diff --git "a/pretrain/trainer_state.json" "b/pretrain/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/pretrain/trainer_state.json"
@@ -0,0 +1,15309 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2181,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004585052728106373,
+      "grad_norm": 82.61045837402344,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 6.2624,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009170105456212746,
+      "grad_norm": 89.19871520996094,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 6.2103,
+      "step": 2
+    },
+    {
+      "epoch": 0.001375515818431912,
+      "grad_norm": 68.57350158691406,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 5.8778,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018340210912425492,
+      "grad_norm": 34.0018310546875,
+      "learning_rate": 6.060606060606061e-05,
+      "loss": 5.4775,
+      "step": 4
+    },
+    {
+      "epoch": 0.0022925263640531865,
+      "grad_norm": 20.510055541992188,
+      "learning_rate": 7.575757575757576e-05,
+      "loss": 5.4113,
+      "step": 5
+    },
+    {
+      "epoch": 0.002751031636863824,
+      "grad_norm": 30.034191131591797,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 5.1566,
+      "step": 6
+    },
+    {
+      "epoch": 0.0032095369096744614,
+      "grad_norm": 9.851234436035156,
+      "learning_rate": 0.00010606060606060606,
+      "loss": 5.1846,
+      "step": 7
+    },
+    {
+      "epoch": 0.0036680421824850985,
+      "grad_norm": 12.711838722229004,
+      "learning_rate": 0.00012121212121212122,
+      "loss": 4.9472,
+      "step": 8
+    },
+    {
+      "epoch": 0.0041265474552957355,
+      "grad_norm": 12.295337677001953,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 4.8711,
+      "step": 9
+    },
+    {
+      "epoch": 0.004585052728106373,
+      "grad_norm": 6.2435736656188965,
+      "learning_rate": 0.00015151515151515152,
+      "loss": 4.6397,
+      "step": 10
+    },
+    {
+      "epoch": 0.0050435580009170105,
+      "grad_norm": 5.533956050872803,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 4.4397,
+      "step": 11
+    },
+    {
+      "epoch": 0.005502063273727648,
+      "grad_norm": 5.272793292999268,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 4.4639,
+      "step": 12
+    },
+    {
+      "epoch": 0.005960568546538285,
+      "grad_norm": 5.0414137840271,
+      "learning_rate": 0.00019696969696969695,
+      "loss": 4.2985,
+      "step": 13
+    },
+    {
+      "epoch": 0.006419073819348923,
+      "grad_norm": 3.2253410816192627,
+      "learning_rate": 0.00021212121212121213,
+      "loss": 4.1735,
+      "step": 14
+    },
+    {
+      "epoch": 0.0068775790921595595,
+      "grad_norm": 3.7403910160064697,
+      "learning_rate": 0.00022727272727272727,
+      "loss": 4.037,
+      "step": 15
+    },
+    {
+      "epoch": 0.007336084364970197,
+      "grad_norm": 2.593864917755127,
+      "learning_rate": 0.00024242424242424245,
+      "loss": 4.0309,
+      "step": 16
+    },
+    {
+      "epoch": 0.007794589637780834,
+      "grad_norm": 2.3420021533966064,
+      "learning_rate": 0.00025757575757575756,
+      "loss": 3.9498,
+      "step": 17
+    },
+    {
+      "epoch": 0.008253094910591471,
+      "grad_norm": 1.9448741674423218,
+      "learning_rate": 0.00027272727272727274,
+      "loss": 3.8132,
+      "step": 18
+    },
+    {
+      "epoch": 0.00871160018340211,
+      "grad_norm": 1.801308035850525,
+      "learning_rate": 0.0002878787878787879,
+      "loss": 3.7884,
+      "step": 19
+    },
+    {
+      "epoch": 0.009170105456212746,
+      "grad_norm": 1.3565490245819092,
+      "learning_rate": 0.00030303030303030303,
+      "loss": 3.8866,
+      "step": 20
+    },
+    {
+      "epoch": 0.009628610729023384,
+      "grad_norm": 1.1786794662475586,
+      "learning_rate": 0.0003181818181818182,
+      "loss": 3.8682,
+      "step": 21
+    },
+    {
+      "epoch": 0.010087116001834021,
+      "grad_norm": 1.1641144752502441,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 3.7139,
+      "step": 22
+    },
+    {
+      "epoch": 0.01054562127464466,
+      "grad_norm": 0.9681243896484375,
+      "learning_rate": 0.0003484848484848485,
+      "loss": 3.6648,
+      "step": 23
+    },
+    {
+      "epoch": 0.011004126547455296,
+      "grad_norm": 1.098687767982483,
+      "learning_rate": 0.00036363636363636367,
+      "loss": 3.7056,
+      "step": 24
+    },
+    {
+      "epoch": 0.011462631820265932,
+      "grad_norm": 1.214459776878357,
+      "learning_rate": 0.0003787878787878788,
+      "loss": 3.671,
+      "step": 25
+    },
+    {
+      "epoch": 0.01192113709307657,
+      "grad_norm": 1.090802788734436,
+      "learning_rate": 0.0003939393939393939,
+      "loss": 3.6303,
+      "step": 26
+    },
+    {
+      "epoch": 0.012379642365887207,
+      "grad_norm": 0.968255877494812,
+      "learning_rate": 0.00040909090909090913,
+      "loss": 3.6686,
+      "step": 27
+    },
+    {
+      "epoch": 0.012838147638697846,
+      "grad_norm": 1.0167286396026611,
+      "learning_rate": 0.00042424242424242425,
+      "loss": 3.5451,
+      "step": 28
+    },
+    {
+      "epoch": 0.013296652911508482,
+      "grad_norm": 1.0673296451568604,
+      "learning_rate": 0.0004393939393939394,
+      "loss": 3.6704,
+      "step": 29
+    },
+    {
+      "epoch": 0.013755158184319119,
+      "grad_norm": 0.780281126499176,
+      "learning_rate": 0.00045454545454545455,
+      "loss": 3.5425,
+      "step": 30
+    },
+    {
+      "epoch": 0.014213663457129757,
+      "grad_norm": 1.0210065841674805,
+      "learning_rate": 0.0004696969696969697,
+      "loss": 3.565,
+      "step": 31
+    },
+    {
+      "epoch": 0.014672168729940394,
+      "grad_norm": 0.6757476329803467,
+      "learning_rate": 0.0004848484848484849,
+      "loss": 3.5947,
+      "step": 32
+    },
+    {
+      "epoch": 0.015130674002751032,
+      "grad_norm": 0.6970779895782471,
+      "learning_rate": 0.0005,
+      "loss": 3.6019,
+      "step": 33
+    },
+    {
+      "epoch": 0.015589179275561669,
+      "grad_norm": 0.5611505508422852,
+      "learning_rate": 0.0005151515151515151,
+      "loss": 3.5799,
+      "step": 34
+    },
+    {
+      "epoch": 0.016047684548372305,
+      "grad_norm": 0.5632553696632385,
+      "learning_rate": 0.0005303030303030302,
+      "loss": 3.5101,
+      "step": 35
+    },
+    {
+      "epoch": 0.016506189821182942,
+      "grad_norm": 0.6619636416435242,
+      "learning_rate": 0.0005454545454545455,
+      "loss": 3.4713,
+      "step": 36
+    },
+    {
+      "epoch": 0.016964695093993582,
+      "grad_norm": 0.5907987356185913,
+      "learning_rate": 0.0005606060606060606,
+      "loss": 3.5065,
+      "step": 37
+    },
+    {
+      "epoch": 0.01742320036680422,
+      "grad_norm": 0.45250579714775085,
+      "learning_rate": 0.0005757575757575758,
+      "loss": 3.467,
+      "step": 38
+    },
+    {
+      "epoch": 0.017881705639614855,
+      "grad_norm": 0.5025972127914429,
+      "learning_rate": 0.0005909090909090909,
+      "loss": 3.4957,
+      "step": 39
+    },
+    {
+      "epoch": 0.018340210912425492,
+      "grad_norm": 0.4533500671386719,
+      "learning_rate": 0.0006060606060606061,
+      "loss": 3.4778,
+      "step": 40
+    },
+    {
+      "epoch": 0.01879871618523613,
+      "grad_norm": 0.4916984438896179,
+      "learning_rate": 0.0006212121212121212,
+      "loss": 3.3474,
+      "step": 41
+    },
+    {
+      "epoch": 0.01925722145804677,
+      "grad_norm": 0.4780777096748352,
+      "learning_rate": 0.0006363636363636364,
+      "loss": 3.4381,
+      "step": 42
+    },
+    {
+      "epoch": 0.019715726730857405,
+      "grad_norm": 0.4316171109676361,
+      "learning_rate": 0.0006515151515151515,
+      "loss": 3.5278,
+      "step": 43
+    },
+    {
+      "epoch": 0.020174232003668042,
+      "grad_norm": 0.43871739506721497,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 3.3684,
+      "step": 44
+    },
+    {
+      "epoch": 0.02063273727647868,
+      "grad_norm": 0.4260846972465515,
+      "learning_rate": 0.0006818181818181818,
+      "loss": 3.3597,
+      "step": 45
+    },
+    {
+      "epoch": 0.02109124254928932,
+      "grad_norm": 0.38816410303115845,
+      "learning_rate": 0.000696969696969697,
+      "loss": 3.4473,
+      "step": 46
+    },
+    {
+      "epoch": 0.021549747822099955,
+      "grad_norm": 0.4049800932407379,
+      "learning_rate": 0.0007121212121212122,
+      "loss": 3.3733,
+      "step": 47
+    },
+    {
+      "epoch": 0.02200825309491059,
+      "grad_norm": 0.4236099421977997,
+      "learning_rate": 0.0007272727272727273,
+      "loss": 3.3447,
+      "step": 48
+    },
+    {
+      "epoch": 0.02246675836772123,
+      "grad_norm": 0.3743979334831238,
+      "learning_rate": 0.0007424242424242425,
+      "loss": 3.5118,
+      "step": 49
+    },
+    {
+      "epoch": 0.022925263640531865,
+      "grad_norm": 0.43707674741744995,
+      "learning_rate": 0.0007575757575757576,
+      "loss": 3.4029,
+      "step": 50
+    },
+    {
+      "epoch": 0.023383768913342505,
+      "grad_norm": 0.38097211718559265,
+      "learning_rate": 0.0007727272727272727,
+      "loss": 3.4068,
+      "step": 51
+    },
+    {
+      "epoch": 0.02384227418615314,
+      "grad_norm": 0.3854580819606781,
+      "learning_rate": 0.0007878787878787878,
+      "loss": 3.4796,
+      "step": 52
+    },
+    {
+      "epoch": 0.024300779458963778,
+      "grad_norm": 0.46823471784591675,
+      "learning_rate": 0.000803030303030303,
+      "loss": 3.4659,
+      "step": 53
+    },
+    {
+      "epoch": 0.024759284731774415,
+      "grad_norm": 0.4682977497577667,
+      "learning_rate": 0.0008181818181818183,
+      "loss": 3.4008,
+      "step": 54
+    },
+    {
+      "epoch": 0.02521779000458505,
+      "grad_norm": 0.42602428793907166,
+      "learning_rate": 0.0008333333333333334,
+      "loss": 3.424,
+      "step": 55
+    },
+    {
+      "epoch": 0.02567629527739569,
+      "grad_norm": 0.3985247313976288,
+      "learning_rate": 0.0008484848484848485,
+      "loss": 3.4128,
+      "step": 56
+    },
+    {
+      "epoch": 0.026134800550206328,
+      "grad_norm": 0.4055021405220032,
+      "learning_rate": 0.0008636363636363636,
+      "loss": 3.4062,
+      "step": 57
+    },
+    {
+      "epoch": 0.026593305823016965,
+      "grad_norm": 0.5106348395347595,
+      "learning_rate": 0.0008787878787878789,
+      "loss": 3.431,
+      "step": 58
+    },
+    {
+      "epoch": 0.0270518110958276,
+      "grad_norm": 0.4446623623371124,
+      "learning_rate": 0.000893939393939394,
+      "loss": 3.3587,
+      "step": 59
+    },
+    {
+      "epoch": 0.027510316368638238,
+      "grad_norm": 0.4808119535446167,
+      "learning_rate": 0.0009090909090909091,
+      "loss": 3.3335,
+      "step": 60
+    },
+    {
+      "epoch": 0.027968821641448878,
+      "grad_norm": 0.4799236059188843,
+      "learning_rate": 0.0009242424242424242,
+      "loss": 3.2693,
+      "step": 61
+    },
+    {
+      "epoch": 0.028427326914259515,
+      "grad_norm": 0.4533957242965698,
+      "learning_rate": 0.0009393939393939394,
+      "loss": 3.311,
+      "step": 62
+    },
+    {
+      "epoch": 0.02888583218707015,
+      "grad_norm": 0.47885721921920776,
+      "learning_rate": 0.0009545454545454546,
+      "loss": 3.3479,
+      "step": 63
+    },
+    {
+      "epoch": 0.029344337459880788,
+      "grad_norm": 0.5001066327095032,
+      "learning_rate": 0.0009696969696969698,
+      "loss": 3.4795,
+      "step": 64
+    },
+    {
+      "epoch": 0.029802842732691424,
+      "grad_norm": 0.5002239346504211,
+      "learning_rate": 0.000984848484848485,
+      "loss": 3.2833,
+      "step": 65
+    },
+    {
+      "epoch": 0.030261348005502064,
+      "grad_norm": 0.5772806406021118,
+      "learning_rate": 0.001,
+      "loss": 3.3471,
+      "step": 66
+    },
+    {
+      "epoch": 0.0307198532783127,
+      "grad_norm": 0.7666266560554504,
+      "learning_rate": 0.0009999994484067654,
+      "loss": 3.3409,
+      "step": 67
+    },
+    {
+      "epoch": 0.031178358551123338,
+      "grad_norm": 0.5101966261863708,
+      "learning_rate": 0.0009999977936282788,
+      "loss": 3.2757,
+      "step": 68
+    },
+    {
+      "epoch": 0.03163686382393398,
+      "grad_norm": 0.6145191788673401,
+      "learning_rate": 0.0009999950356681913,
+      "loss": 3.3627,
+      "step": 69
+    },
+    {
+      "epoch": 0.03209536909674461,
+      "grad_norm": 0.49831077456474304,
+      "learning_rate": 0.0009999911745325876,
+      "loss": 3.2484,
+      "step": 70
+    },
+    {
+      "epoch": 0.03255387436955525,
+      "grad_norm": 0.6562492847442627,
+      "learning_rate": 0.0009999862102299873,
+      "loss": 3.2895,
+      "step": 71
+    },
+    {
+      "epoch": 0.033012379642365884,
+      "grad_norm": 0.7735105156898499,
+      "learning_rate": 0.0009999801427713433,
+      "loss": 3.3778,
+      "step": 72
+    },
+    {
+      "epoch": 0.033470884915176524,
+      "grad_norm": 0.4456319808959961,
+      "learning_rate": 0.0009999729721700424,
+      "loss": 3.225,
+      "step": 73
+    },
+    {
+      "epoch": 0.033929390187987164,
+      "grad_norm": 0.6353105902671814,
+      "learning_rate": 0.000999964698441906,
+      "loss": 3.2982,
+      "step": 74
+    },
+    {
+      "epoch": 0.0343878954607978,
+      "grad_norm": 0.6268491148948669,
+      "learning_rate": 0.0009999553216051892,
+      "loss": 3.3352,
+      "step": 75
+    },
+    {
+      "epoch": 0.03484640073360844,
+      "grad_norm": 0.5871131420135498,
+      "learning_rate": 0.00099994484168058,
+      "loss": 3.2977,
+      "step": 76
+    },
+    {
+      "epoch": 0.03530490600641907,
+      "grad_norm": 0.6387215852737427,
+      "learning_rate": 0.0009999332586912019,
+      "loss": 3.3409,
+      "step": 77
+    },
+    {
+      "epoch": 0.03576341127922971,
+      "grad_norm": 0.5404582619667053,
+      "learning_rate": 0.0009999205726626108,
+      "loss": 3.275,
+      "step": 78
+    },
+    {
+      "epoch": 0.03622191655204035,
+      "grad_norm": 0.7002769708633423,
+      "learning_rate": 0.000999906783622797,
+      "loss": 3.259,
+      "step": 79
+    },
+    {
+      "epoch": 0.036680421824850984,
+      "grad_norm": 0.7378426790237427,
+      "learning_rate": 0.0009998918916021842,
+      "loss": 3.3242,
+      "step": 80
+    },
+    {
+      "epoch": 0.037138927097661624,
+      "grad_norm": 0.7165883183479309,
+      "learning_rate": 0.0009998758966336297,
+      "loss": 3.2121,
+      "step": 81
+    },
+    {
+      "epoch": 0.03759743237047226,
+      "grad_norm": 0.9018951654434204,
+      "learning_rate": 0.0009998587987524242,
+      "loss": 3.2819,
+      "step": 82
+    },
+    {
+      "epoch": 0.0380559376432829,
+      "grad_norm": 0.8918347954750061,
+      "learning_rate": 0.0009998405979962926,
+      "loss": 3.2762,
+      "step": 83
+    },
+    {
+      "epoch": 0.03851444291609354,
+      "grad_norm": 1.008992075920105,
+      "learning_rate": 0.000999821294405392,
+      "loss": 3.266,
+      "step": 84
+    },
+    {
+      "epoch": 0.03897294818890417,
+      "grad_norm": 0.9068806171417236,
+      "learning_rate": 0.0009998008880223134,
+      "loss": 3.3072,
+      "step": 85
+    },
+    {
+      "epoch": 0.03943145346171481,
+      "grad_norm": 0.7586443424224854,
+      "learning_rate": 0.000999779378892081,
+      "loss": 3.341,
+      "step": 86
+    },
+    {
+      "epoch": 0.039889958734525444,
+      "grad_norm": 0.709721028804779,
+      "learning_rate": 0.0009997567670621522,
+      "loss": 3.1702,
+      "step": 87
+    },
+    {
+      "epoch": 0.040348464007336084,
+      "grad_norm": 0.7303472757339478,
+      "learning_rate": 0.0009997330525824165,
+      "loss": 3.2431,
+      "step": 88
+    },
+    {
+      "epoch": 0.040806969280146724,
+      "grad_norm": 0.7229411602020264,
+      "learning_rate": 0.0009997082355051976,
+      "loss": 3.2379,
+      "step": 89
+    },
+    {
+      "epoch": 0.04126547455295736,
+      "grad_norm": 0.8216807246208191,
+      "learning_rate": 0.000999682315885251,
+      "loss": 3.2529,
+      "step": 90
+    },
+    {
+      "epoch": 0.041723979825768,
+      "grad_norm": 0.8675643801689148,
+      "learning_rate": 0.0009996552937797645,
+      "loss": 3.183,
+      "step": 91
+    },
+    {
+      "epoch": 0.04218248509857864,
+      "grad_norm": 0.8306193351745605,
+      "learning_rate": 0.0009996271692483596,
+      "loss": 3.1675,
+      "step": 92
+    },
+    {
+      "epoch": 0.04264099037138927,
+      "grad_norm": 0.8420382738113403,
+      "learning_rate": 0.0009995979423530893,
+      "loss": 3.1642,
+      "step": 93
+    },
+    {
+      "epoch": 0.04309949564419991,
+      "grad_norm": 1.03010892868042,
+      "learning_rate": 0.000999567613158439,
+      "loss": 3.2202,
+      "step": 94
+    },
+    {
+      "epoch": 0.04355800091701054,
+      "grad_norm": 0.6258354187011719,
+      "learning_rate": 0.0009995361817313263,
+      "loss": 3.2229,
+      "step": 95
+    },
+    {
+      "epoch": 0.04401650618982118,
+      "grad_norm": 0.9931930899620056,
+      "learning_rate": 0.0009995036481411004,
+      "loss": 3.2822,
+      "step": 96
+    },
+    {
+      "epoch": 0.044475011462631824,
+      "grad_norm": 0.8666135668754578,
+      "learning_rate": 0.0009994700124595429,
+      "loss": 3.2227,
+      "step": 97
+    },
+    {
+      "epoch": 0.04493351673544246,
+      "grad_norm": 0.8610184788703918,
+      "learning_rate": 0.0009994352747608663,
+      "loss": 3.195,
+      "step": 98
+    },
+    {
+      "epoch": 0.0453920220082531,
+      "grad_norm": 0.8566797971725464,
+      "learning_rate": 0.0009993994351217151,
+      "loss": 3.1917,
+      "step": 99
+    },
+    {
+      "epoch": 0.04585052728106373,
+      "grad_norm": 0.9598749876022339,
+      "learning_rate": 0.000999362493621165,
+      "loss": 3.188,
+      "step": 100
+    },
+    {
+      "epoch": 0.04630903255387437,
+      "grad_norm": 0.7008885741233826,
+      "learning_rate": 0.0009993244503407226,
+      "loss": 3.2401,
+      "step": 101
+    },
+    {
+      "epoch": 0.04676753782668501,
+      "grad_norm": 0.7476081848144531,
+      "learning_rate": 0.0009992853053643258,
+      "loss": 3.0911,
+      "step": 102
+    },
+    {
+      "epoch": 0.04722604309949564,
+      "grad_norm": 0.6519758105278015,
+      "learning_rate": 0.0009992450587783426,
+      "loss": 3.1783,
+      "step": 103
+    },
+    {
+      "epoch": 0.04768454837230628,
+      "grad_norm": 0.6671428084373474,
+      "learning_rate": 0.000999203710671572,
+      "loss": 3.2377,
+      "step": 104
+    },
+    {
+      "epoch": 0.048143053645116916,
+      "grad_norm": 0.7099413871765137,
+      "learning_rate": 0.0009991612611352438,
+      "loss": 3.1355,
+      "step": 105
+    },
+    {
+      "epoch": 0.048601558917927556,
+      "grad_norm": 0.7780017852783203,
+      "learning_rate": 0.0009991177102630173,
+      "loss": 3.1951,
+      "step": 106
+    },
+    {
+      "epoch": 0.049060064190738197,
+      "grad_norm": 1.0008108615875244,
+      "learning_rate": 0.0009990730581509817,
+      "loss": 3.0903,
+      "step": 107
+    },
+    {
+      "epoch": 0.04951856946354883,
+      "grad_norm": 1.1164809465408325,
+      "learning_rate": 0.0009990273048976566,
+      "loss": 3.088,
+      "step": 108
+    },
+    {
+      "epoch": 0.04997707473635947,
+      "grad_norm": 0.7636913061141968,
+      "learning_rate": 0.0009989804506039905,
+      "loss": 3.1279,
+      "step": 109
+    },
+    {
+      "epoch": 0.0504355800091701,
+      "grad_norm": 0.8368927836418152,
+      "learning_rate": 0.0009989324953733614,
+      "loss": 3.2188,
+      "step": 110
+    },
+    {
+      "epoch": 0.05089408528198074,
+      "grad_norm": 0.7319849133491516,
+      "learning_rate": 0.0009988834393115766,
+      "loss": 3.1851,
+      "step": 111
+    },
+    {
+      "epoch": 0.05135259055479138,
+      "grad_norm": 0.7636634111404419,
+      "learning_rate": 0.000998833282526872,
+      "loss": 3.1016,
+      "step": 112
+    },
+    {
+      "epoch": 0.051811095827602016,
+      "grad_norm": 0.7459844350814819,
+      "learning_rate": 0.0009987820251299122,
+      "loss": 3.062,
+      "step": 113
+    },
+    {
+      "epoch": 0.052269601100412656,
+      "grad_norm": 0.8797902464866638,
+      "learning_rate": 0.00099872966723379,
+      "loss": 3.172,
+      "step": 114
+    },
+    {
+      "epoch": 0.05272810637322329,
+      "grad_norm": 0.741369903087616,
+      "learning_rate": 0.0009986762089540266,
+      "loss": 3.126,
+      "step": 115
+    },
+    {
+      "epoch": 0.05318661164603393,
+      "grad_norm": 0.7959185242652893,
+      "learning_rate": 0.0009986216504085709,
+      "loss": 3.1682,
+      "step": 116
+    },
+    {
+      "epoch": 0.05364511691884457,
+      "grad_norm": 0.8000556230545044,
+      "learning_rate": 0.0009985659917177991,
+      "loss": 3.1948,
+      "step": 117
+    },
+    {
+      "epoch": 0.0541036221916552,
+      "grad_norm": 0.7374656200408936,
+      "learning_rate": 0.0009985092330045155,
+      "loss": 3.0755,
+      "step": 118
+    },
+    {
+      "epoch": 0.05456212746446584,
+      "grad_norm": 0.7770431041717529,
+      "learning_rate": 0.0009984513743939508,
+      "loss": 3.1161,
+      "step": 119
+    },
+    {
+      "epoch": 0.055020632737276476,
+      "grad_norm": 0.7298172116279602,
+      "learning_rate": 0.0009983924160137626,
+      "loss": 3.0882,
+      "step": 120
+    },
+    {
+      "epoch": 0.055479138010087116,
+      "grad_norm": 0.922307014465332,
+      "learning_rate": 0.000998332357994035,
+      "loss": 3.096,
+      "step": 121
+    },
+    {
+      "epoch": 0.055937643282897756,
+      "grad_norm": 0.7248730659484863,
+      "learning_rate": 0.0009982712004672786,
+      "loss": 3.0598,
+      "step": 122
+    },
+    {
+      "epoch": 0.05639614855570839,
+      "grad_norm": 0.6635040044784546,
+      "learning_rate": 0.0009982089435684295,
+      "loss": 3.0949,
+      "step": 123
+    },
+    {
+      "epoch": 0.05685465382851903,
+      "grad_norm": 0.8345446586608887,
+      "learning_rate": 0.0009981455874348499,
+      "loss": 3.0973,
+      "step": 124
+    },
+    {
+      "epoch": 0.05731315910132966,
+      "grad_norm": 0.7962020635604858,
+      "learning_rate": 0.0009980811322063269,
+      "loss": 3.0857,
+      "step": 125
+    },
+    {
+      "epoch": 0.0577716643741403,
+      "grad_norm": 0.9181259274482727,
+      "learning_rate": 0.0009980155780250728,
+      "loss": 3.0782,
+      "step": 126
+    },
+    {
+      "epoch": 0.05823016964695094,
+      "grad_norm": 0.7376343607902527,
+      "learning_rate": 0.0009979489250357243,
+      "loss": 2.9648,
+      "step": 127
+    },
+    {
+      "epoch": 0.058688674919761576,
+      "grad_norm": 0.789767324924469,
+      "learning_rate": 0.0009978811733853431,
+      "loss": 3.0256,
+      "step": 128
+    },
+    {
+      "epoch": 0.059147180192572216,
+      "grad_norm": 0.7572616338729858,
+      "learning_rate": 0.0009978123232234147,
+      "loss": 2.9899,
+      "step": 129
+    },
+    {
+      "epoch": 0.05960568546538285,
+      "grad_norm": 0.6423022747039795,
+      "learning_rate": 0.000997742374701848,
+      "loss": 3.0142,
+      "step": 130
+    },
+    {
+      "epoch": 0.06006419073819349,
+      "grad_norm": 0.7977400422096252,
+      "learning_rate": 0.0009976713279749754,
+      "loss": 2.8752,
+      "step": 131
+    },
+    {
+      "epoch": 0.06052269601100413,
+      "grad_norm": 0.73207688331604,
+      "learning_rate": 0.0009975991831995528,
+      "loss": 3.0459,
+      "step": 132
+    },
+    {
+      "epoch": 0.06098120128381476,
+      "grad_norm": 0.8527759909629822,
+      "learning_rate": 0.0009975259405347581,
+      "loss": 3.0258,
+      "step": 133
+    },
+    {
+      "epoch": 0.0614397065566254,
+      "grad_norm": 0.8217706084251404,
+      "learning_rate": 0.0009974516001421926,
+      "loss": 3.0377,
+      "step": 134
+    },
+    {
+      "epoch": 0.061898211829436035,
+      "grad_norm": 0.8935520052909851,
+      "learning_rate": 0.000997376162185878,
+      "loss": 3.074,
+      "step": 135
+    },
+    {
+      "epoch": 0.062356717102246675,
+      "grad_norm": 0.8200856447219849,
+      "learning_rate": 0.0009972996268322594,
+      "loss": 3.0325,
+      "step": 136
+    },
+    {
+      "epoch": 0.06281522237505731,
+      "grad_norm": 0.84970623254776,
+      "learning_rate": 0.0009972219942502017,
+      "loss": 3.0275,
+      "step": 137
+    },
+    {
+      "epoch": 0.06327372764786796,
+      "grad_norm": 0.8437201976776123,
+      "learning_rate": 0.0009971432646109918,
+      "loss": 3.0069,
+      "step": 138
+    },
+    {
+      "epoch": 0.06373223292067859,
+      "grad_norm": 0.786875307559967,
+      "learning_rate": 0.0009970634380883365,
+      "loss": 2.8974,
+      "step": 139
+    },
+    {
+      "epoch": 0.06419073819348922,
+      "grad_norm": 0.786919116973877,
+      "learning_rate": 0.0009969825148583627,
+      "loss": 2.977,
+      "step": 140
+    },
+    {
+      "epoch": 0.06464924346629987,
+      "grad_norm": 0.8005730509757996,
+      "learning_rate": 0.0009969004950996173,
+      "loss": 2.9512,
+      "step": 141
+    },
+    {
+      "epoch": 0.0651077487391105,
+      "grad_norm": 0.9246258735656738,
+      "learning_rate": 0.0009968173789930668,
+      "loss": 3.0517,
+      "step": 142
+    },
+    {
+      "epoch": 0.06556625401192114,
+      "grad_norm": 0.8451228141784668,
+      "learning_rate": 0.0009967331667220958,
+      "loss": 2.9716,
+      "step": 143
+    },
+    {
+      "epoch": 0.06602475928473177,
+      "grad_norm": 0.955964207649231,
+      "learning_rate": 0.0009966478584725086,
+      "loss": 2.9927,
+      "step": 144
+    },
+    {
+      "epoch": 0.06648326455754242,
+      "grad_norm": 0.7643736600875854,
+      "learning_rate": 0.0009965614544325263,
+      "loss": 2.9932,
+      "step": 145
+    },
+    {
+      "epoch": 0.06694176983035305,
+      "grad_norm": 0.814987063407898,
+      "learning_rate": 0.000996473954792789,
+      "loss": 2.9738,
+      "step": 146
+    },
+    {
+      "epoch": 0.06740027510316368,
+      "grad_norm": 0.7710995078086853,
+      "learning_rate": 0.0009963853597463532,
+      "loss": 2.9742,
+      "step": 147
+    },
+    {
+      "epoch": 0.06785878037597433,
+      "grad_norm": 0.7728829979896545,
+      "learning_rate": 0.000996295669488693,
+      "loss": 3.0066,
+      "step": 148
+    },
+    {
+      "epoch": 0.06831728564878496,
+      "grad_norm": 0.8810952305793762,
+      "learning_rate": 0.0009962048842176979,
+      "loss": 3.0397,
+      "step": 149
+    },
+    {
+      "epoch": 0.0687757909215956,
+      "grad_norm": 0.7474966645240784,
+      "learning_rate": 0.0009961130041336748,
+      "loss": 2.9172,
+      "step": 150
+    },
+    {
+      "epoch": 0.06923429619440624,
+      "grad_norm": 0.922071099281311,
+      "learning_rate": 0.0009960200294393449,
+      "loss": 2.9832,
+      "step": 151
+    },
+    {
+      "epoch": 0.06969280146721687,
+      "grad_norm": 0.7933943271636963,
+      "learning_rate": 0.0009959259603398453,
+      "loss": 2.926,
+      "step": 152
+    },
+    {
+      "epoch": 0.07015130674002751,
+      "grad_norm": 0.981529951095581,
+      "learning_rate": 0.0009958307970427275,
+      "loss": 2.8658,
+      "step": 153
+    },
+    {
+      "epoch": 0.07060981201283814,
+      "grad_norm": 0.8337743878364563,
+      "learning_rate": 0.0009957345397579572,
+      "loss": 2.951,
+      "step": 154
+    },
+    {
+      "epoch": 0.07106831728564879,
+      "grad_norm": 0.9412785768508911,
+      "learning_rate": 0.0009956371886979138,
+      "loss": 2.9722,
+      "step": 155
+    },
+    {
+      "epoch": 0.07152682255845942,
+      "grad_norm": 0.8368416428565979,
+      "learning_rate": 0.00099553874407739,
+      "loss": 2.952,
+      "step": 156
+    },
+    {
+      "epoch": 0.07198532783127005,
+      "grad_norm": 0.7955209612846375,
+      "learning_rate": 0.0009954392061135916,
+      "loss": 2.9508,
+      "step": 157
+    },
+    {
+      "epoch": 0.0724438331040807,
+      "grad_norm": 0.7679097652435303,
+      "learning_rate": 0.0009953385750261364,
+      "loss": 2.9491,
+      "step": 158
+    },
+    {
+      "epoch": 0.07290233837689133,
+      "grad_norm": 0.7876397967338562,
+      "learning_rate": 0.0009952368510370538,
+      "loss": 2.9624,
+      "step": 159
+    },
+    {
+      "epoch": 0.07336084364970197,
+      "grad_norm": 0.9138140678405762,
+      "learning_rate": 0.0009951340343707852,
+      "loss": 2.9538,
+      "step": 160
+    },
+    {
+      "epoch": 0.07381934892251261,
+      "grad_norm": 0.9184808135032654,
+      "learning_rate": 0.0009950301252541823,
+      "loss": 2.9179,
+      "step": 161
+    },
+    {
+      "epoch": 0.07427785419532325,
+      "grad_norm": 0.9548496603965759,
+      "learning_rate": 0.0009949251239165075,
+      "loss": 2.873,
+      "step": 162
+    },
+    {
+      "epoch": 0.07473635946813388,
+      "grad_norm": 0.9101412296295166,
+      "learning_rate": 0.000994819030589433,
+      "loss": 2.9607,
+      "step": 163
+    },
+    {
+      "epoch": 0.07519486474094451,
+      "grad_norm": 0.8322324156761169,
+      "learning_rate": 0.00099471184550704,
+      "loss": 2.8869,
+      "step": 164
+    },
+    {
+      "epoch": 0.07565337001375516,
+      "grad_norm": 0.9734076857566833,
+      "learning_rate": 0.0009946035689058189,
+      "loss": 2.8981,
+      "step": 165
+    },
+    {
+      "epoch": 0.0761118752865658,
+      "grad_norm": 0.9089769124984741,
+      "learning_rate": 0.0009944942010246681,
+      "loss": 2.9517,
+      "step": 166
+    },
+    {
+      "epoch": 0.07657038055937643,
+      "grad_norm": 0.9945159554481506,
+      "learning_rate": 0.0009943837421048942,
+      "loss": 2.97,
+      "step": 167
+    },
+    {
+      "epoch": 0.07702888583218707,
+      "grad_norm": 0.9766585230827332,
+      "learning_rate": 0.0009942721923902106,
+      "loss": 2.8009,
+      "step": 168
+    },
+    {
+      "epoch": 0.07748739110499771,
+      "grad_norm": 0.9825783371925354,
+      "learning_rate": 0.0009941595521267377,
+      "loss": 2.7362,
+      "step": 169
+    },
+    {
+      "epoch": 0.07794589637780834,
+      "grad_norm": 1.1186515092849731,
+      "learning_rate": 0.0009940458215630017,
+      "loss": 2.8868,
+      "step": 170
+    },
+    {
+      "epoch": 0.07840440165061899,
+      "grad_norm": 0.9937232136726379,
+      "learning_rate": 0.0009939310009499348,
+      "loss": 2.9241,
+      "step": 171
+    },
+    {
+      "epoch": 0.07886290692342962,
+      "grad_norm": 1.2268695831298828,
+      "learning_rate": 0.000993815090540874,
+      "loss": 2.9258,
+      "step": 172
+    },
+    {
+      "epoch": 0.07932141219624025,
+      "grad_norm": 0.7727352976799011,
+      "learning_rate": 0.000993698090591561,
+      "loss": 2.948,
+      "step": 173
+    },
+    {
+      "epoch": 0.07977991746905089,
+      "grad_norm": 0.9068551063537598,
+      "learning_rate": 0.0009935800013601416,
+      "loss": 2.8249,
+      "step": 174
+    },
+    {
+      "epoch": 0.08023842274186153,
+      "grad_norm": 0.9239626526832581,
+      "learning_rate": 0.000993460823107164,
+      "loss": 2.8257,
+      "step": 175
+    },
+    {
+      "epoch": 0.08069692801467217,
+      "grad_norm": 0.7615908980369568,
+      "learning_rate": 0.0009933405560955803,
+      "loss": 2.837,
+      "step": 176
+    },
+    {
+      "epoch": 0.0811554332874828,
+      "grad_norm": 0.9969072937965393,
+      "learning_rate": 0.0009932192005907446,
+      "loss": 2.8473,
+      "step": 177
+    },
+    {
+      "epoch": 0.08161393856029345,
+      "grad_norm": 0.8133277297019958,
+      "learning_rate": 0.0009930967568604118,
+      "loss": 2.9116,
+      "step": 178
+    },
+    {
+      "epoch": 0.08207244383310408,
+      "grad_norm": 1.0376125574111938,
+      "learning_rate": 0.000992973225174739,
+      "loss": 2.8204,
+      "step": 179
+    },
+    {
+      "epoch": 0.08253094910591471,
+      "grad_norm": 0.8388999700546265,
+      "learning_rate": 0.0009928486058062827,
+      "loss": 2.8268,
+      "step": 180
+    },
+    {
+      "epoch": 0.08298945437872536,
+      "grad_norm": 0.776506245136261,
+      "learning_rate": 0.0009927228990299999,
+      "loss": 2.8918,
+      "step": 181
+    },
+    {
+      "epoch": 0.083447959651536,
+      "grad_norm": 0.8521844744682312,
+      "learning_rate": 0.0009925961051232468,
+      "loss": 2.8517,
+      "step": 182
+    },
+    {
+      "epoch": 0.08390646492434663,
+      "grad_norm": 0.8654164671897888,
+      "learning_rate": 0.000992468224365778,
+      "loss": 2.7635,
+      "step": 183
+    },
+    {
+      "epoch": 0.08436497019715727,
+      "grad_norm": 0.8541561365127563,
+      "learning_rate": 0.000992339257039746,
+      "loss": 2.8239,
+      "step": 184
+    },
+    {
+      "epoch": 0.08482347546996791,
+      "grad_norm": 0.7930411100387573,
+      "learning_rate": 0.0009922092034297006,
+      "loss": 2.7288,
+      "step": 185
+    },
+    {
+      "epoch": 0.08528198074277854,
+      "grad_norm": 0.9656765460968018,
+      "learning_rate": 0.0009920780638225891,
+      "loss": 2.7736,
+      "step": 186
+    },
+    {
+      "epoch": 0.08574048601558917,
+      "grad_norm": 0.8736869096755981,
+      "learning_rate": 0.0009919458385077538,
+      "loss": 2.7799,
+      "step": 187
+    },
+    {
+      "epoch": 0.08619899128839982,
+      "grad_norm": 0.9572360515594482,
+      "learning_rate": 0.0009918125277769336,
+      "loss": 2.7708,
+      "step": 188
+    },
+    {
+      "epoch": 0.08665749656121045,
+      "grad_norm": 0.908798336982727,
+      "learning_rate": 0.0009916781319242614,
+      "loss": 2.8481,
+      "step": 189
+    },
+    {
+      "epoch": 0.08711600183402109,
+      "grad_norm": 0.8614920973777771,
+      "learning_rate": 0.0009915426512462646,
+      "loss": 2.8141,
+      "step": 190
+    },
+    {
+      "epoch": 0.08757450710683173,
+      "grad_norm": 0.9561100006103516,
+      "learning_rate": 0.0009914060860418644,
+      "loss": 2.837,
+      "step": 191
+    },
+    {
+      "epoch": 0.08803301237964237,
+      "grad_norm": 0.8394006490707397,
+      "learning_rate": 0.000991268436612374,
+      "loss": 2.828,
+      "step": 192
+    },
+    {
+      "epoch": 0.088491517652453,
+      "grad_norm": 0.8883221745491028,
+      "learning_rate": 0.0009911297032614997,
+      "loss": 2.8347,
+      "step": 193
+    },
+    {
+      "epoch": 0.08895002292526365,
+      "grad_norm": 0.9614878296852112,
+      "learning_rate": 0.000990989886295339,
+      "loss": 2.8076,
+      "step": 194
+    },
+    {
+      "epoch": 0.08940852819807428,
+      "grad_norm": 0.9566423892974854,
+      "learning_rate": 0.0009908489860223804,
+      "loss": 2.7599,
+      "step": 195
+    },
+    {
+      "epoch": 0.08986703347088491,
+      "grad_norm": 0.9820214509963989,
+      "learning_rate": 0.000990707002753502,
+      "loss": 2.7882,
+      "step": 196
+    },
+    {
+      "epoch": 0.09032553874369555,
+      "grad_norm": 0.9029565453529358,
+      "learning_rate": 0.0009905639368019724,
+      "loss": 2.7489,
+      "step": 197
+    },
+    {
+      "epoch": 0.0907840440165062,
+      "grad_norm": 0.8777222037315369,
+      "learning_rate": 0.0009904197884834482,
+      "loss": 2.7625,
+      "step": 198
+    },
+    {
+      "epoch": 0.09124254928931683,
+      "grad_norm": 0.9996369481086731,
+      "learning_rate": 0.0009902745581159742,
+      "loss": 2.7772,
+      "step": 199
+    },
+    {
+      "epoch": 0.09170105456212746,
+      "grad_norm": 1.100584864616394,
+      "learning_rate": 0.0009901282460199829,
+      "loss": 2.777,
+      "step": 200
+    },
+    {
+      "epoch": 0.0921595598349381,
+      "grad_norm": 0.9204021096229553,
+      "learning_rate": 0.0009899808525182935,
+      "loss": 2.7303,
+      "step": 201
+    },
+    {
+      "epoch": 0.09261806510774874,
+      "grad_norm": 0.9471940398216248,
+      "learning_rate": 0.0009898323779361107,
+      "loss": 2.6648,
+      "step": 202
+    },
+    {
+      "epoch": 0.09307657038055937,
+      "grad_norm": 0.9402942061424255,
+      "learning_rate": 0.000989682822601025,
+      "loss": 2.7532,
+      "step": 203
+    },
+    {
+      "epoch": 0.09353507565337002,
+      "grad_norm": 1.0100222826004028,
+      "learning_rate": 0.0009895321868430113,
+      "loss": 2.7108,
+      "step": 204
+    },
+    {
+      "epoch": 0.09399358092618065,
+      "grad_norm": 0.967642068862915,
+      "learning_rate": 0.0009893804709944281,
+      "loss": 2.7457,
+      "step": 205
+    },
+    {
+      "epoch": 0.09445208619899129,
+      "grad_norm": 1.0845133066177368,
+      "learning_rate": 0.0009892276753900174,
+      "loss": 2.7493,
+      "step": 206
+    },
+    {
+      "epoch": 0.09491059147180192,
+      "grad_norm": 1.018634557723999,
+      "learning_rate": 0.0009890738003669028,
+      "loss": 2.7248,
+      "step": 207
+    },
+    {
+      "epoch": 0.09536909674461257,
+      "grad_norm": 1.016206979751587,
+      "learning_rate": 0.0009889188462645904,
+      "loss": 2.7687,
+      "step": 208
+    },
+    {
+      "epoch": 0.0958276020174232,
+      "grad_norm": 0.9672858715057373,
+      "learning_rate": 0.0009887628134249667,
+      "loss": 2.7425,
+      "step": 209
+    },
+    {
+      "epoch": 0.09628610729023383,
+      "grad_norm": 1.0289807319641113,
+      "learning_rate": 0.0009886057021922983,
+      "loss": 2.7908,
+      "step": 210
+    },
+    {
+      "epoch": 0.09674461256304448,
+      "grad_norm": 1.1172319650650024,
+      "learning_rate": 0.0009884475129132311,
+      "loss": 2.723,
+      "step": 211
+    },
+    {
+      "epoch": 0.09720311783585511,
+      "grad_norm": 1.2268468141555786,
+      "learning_rate": 0.0009882882459367897,
+      "loss": 2.6337,
+      "step": 212
+    },
+    {
+      "epoch": 0.09766162310866575,
+      "grad_norm": 0.9319258332252502,
+      "learning_rate": 0.0009881279016143766,
+      "loss": 2.5076,
+      "step": 213
+    },
+    {
+      "epoch": 0.09812012838147639,
+      "grad_norm": 1.0663557052612305,
+      "learning_rate": 0.0009879664802997707,
+      "loss": 2.7519,
+      "step": 214
+    },
+    {
+      "epoch": 0.09857863365428703,
+      "grad_norm": 0.9507728815078735,
+      "learning_rate": 0.000987803982349128,
+      "loss": 2.7308,
+      "step": 215
+    },
+    {
+      "epoch": 0.09903713892709766,
+      "grad_norm": 0.9936598539352417,
+      "learning_rate": 0.0009876404081209796,
+      "loss": 2.7146,
+      "step": 216
+    },
+    {
+      "epoch": 0.09949564419990829,
+      "grad_norm": 1.0016199350357056,
+      "learning_rate": 0.000987475757976231,
+      "loss": 2.7453,
+      "step": 217
+    },
+    {
+      "epoch": 0.09995414947271894,
+      "grad_norm": 0.9770188331604004,
+      "learning_rate": 0.000987310032278162,
+      "loss": 2.6561,
+      "step": 218
+    },
+    {
+      "epoch": 0.10041265474552957,
+      "grad_norm": 0.9570183753967285,
+      "learning_rate": 0.0009871432313924254,
+      "loss": 2.6405,
+      "step": 219
+    },
+    {
+      "epoch": 0.1008711600183402,
+      "grad_norm": 0.9441530108451843,
+      "learning_rate": 0.000986975355687046,
+      "loss": 2.6735,
+      "step": 220
+    },
+    {
+      "epoch": 0.10132966529115085,
+      "grad_norm": 1.0475291013717651,
+      "learning_rate": 0.0009868064055324204,
+      "loss": 2.6198,
+      "step": 221
+    },
+    {
+      "epoch": 0.10178817056396149,
+      "grad_norm": 1.0661554336547852,
+      "learning_rate": 0.0009866363813013153,
+      "loss": 2.6608,
+      "step": 222
+    },
+    {
+      "epoch": 0.10224667583677212,
+      "grad_norm": 0.9617491364479065,
+      "learning_rate": 0.0009864652833688676,
+      "loss": 2.6339,
+      "step": 223
+    },
+    {
+      "epoch": 0.10270518110958277,
+      "grad_norm": 0.9131588339805603,
+      "learning_rate": 0.0009862931121125836,
+      "loss": 2.6225,
+      "step": 224
+    },
+    {
+      "epoch": 0.1031636863823934,
+      "grad_norm": 1.1531404256820679,
+      "learning_rate": 0.000986119867912337,
+      "loss": 2.6611,
+      "step": 225
+    },
+    {
+      "epoch": 0.10362219165520403,
+      "grad_norm": 0.9137834906578064,
+      "learning_rate": 0.000985945551150369,
+      "loss": 2.6499,
+      "step": 226
+    },
+    {
+      "epoch": 0.10408069692801467,
+      "grad_norm": 0.8744311928749084,
+      "learning_rate": 0.0009857701622112876,
+      "loss": 2.5614,
+      "step": 227
+    },
+    {
+      "epoch": 0.10453920220082531,
+      "grad_norm": 0.8695749640464783,
+      "learning_rate": 0.000985593701482066,
+      "loss": 2.6878,
+      "step": 228
+    },
+    {
+      "epoch": 0.10499770747363595,
+      "grad_norm": 0.8920214176177979,
+      "learning_rate": 0.0009854161693520424,
+      "loss": 2.619,
+      "step": 229
+    },
+    {
+      "epoch": 0.10545621274644658,
+      "grad_norm": 0.9327085614204407,
+      "learning_rate": 0.0009852375662129194,
+      "loss": 2.6288,
+      "step": 230
+    },
+    {
+      "epoch": 0.10591471801925723,
+      "grad_norm": 0.9928507208824158,
+      "learning_rate": 0.0009850578924587613,
+      "loss": 2.5572,
+      "step": 231
+    },
+    {
+      "epoch": 0.10637322329206786,
+      "grad_norm": 0.8341806530952454,
+      "learning_rate": 0.000984877148485996,
+      "loss": 2.6321,
+      "step": 232
+    },
+    {
+      "epoch": 0.10683172856487849,
+      "grad_norm": 1.0217528343200684,
+      "learning_rate": 0.000984695334693412,
+      "loss": 2.6769,
+      "step": 233
+    },
+    {
+      "epoch": 0.10729023383768914,
+      "grad_norm": 0.9497523307800293,
+      "learning_rate": 0.000984512451482158,
+      "loss": 2.6142,
+      "step": 234
+    },
+    {
+      "epoch": 0.10774873911049977,
+      "grad_norm": 0.92038893699646,
+      "learning_rate": 0.0009843284992557431,
+      "loss": 2.6078,
+      "step": 235
+    },
+    {
+      "epoch": 0.1082072443833104,
+      "grad_norm": 0.8683038353919983,
+      "learning_rate": 0.000984143478420034,
+      "loss": 2.6432,
+      "step": 236
+    },
+    {
+      "epoch": 0.10866574965612105,
+      "grad_norm": 0.8895754814147949,
+      "learning_rate": 0.0009839573893832563,
+      "loss": 2.6071,
+      "step": 237
+    },
+    {
+      "epoch": 0.10912425492893169,
+      "grad_norm": 0.8092901110649109,
+      "learning_rate": 0.000983770232555991,
+      "loss": 2.681,
+      "step": 238
+    },
+    {
+      "epoch": 0.10958276020174232,
+      "grad_norm": 0.9066882729530334,
+      "learning_rate": 0.0009835820083511765,
+      "loss": 2.5356,
+      "step": 239
+    },
+    {
+      "epoch": 0.11004126547455295,
+      "grad_norm": 0.8102091550827026,
+      "learning_rate": 0.0009833927171841055,
+      "loss": 2.6425,
+      "step": 240
+    },
+    {
+      "epoch": 0.1104997707473636,
+      "grad_norm": 1.0509649515151978,
+      "learning_rate": 0.0009832023594724246,
+      "loss": 2.636,
+      "step": 241
+    },
+    {
+      "epoch": 0.11095827602017423,
+      "grad_norm": 0.9189571142196655,
+      "learning_rate": 0.0009830109356361344,
+      "loss": 2.5905,
+      "step": 242
+    },
+    {
+      "epoch": 0.11141678129298486,
+      "grad_norm": 0.8751232028007507,
+      "learning_rate": 0.0009828184460975867,
+      "loss": 2.5595,
+      "step": 243
+    },
+    {
+      "epoch": 0.11187528656579551,
+      "grad_norm": 0.7567734718322754,
+      "learning_rate": 0.0009826248912814855,
+      "loss": 2.5015,
+      "step": 244
+    },
+    {
+      "epoch": 0.11233379183860615,
+      "grad_norm": 0.8169068098068237,
+      "learning_rate": 0.0009824302716148847,
+      "loss": 2.5973,
+      "step": 245
+    },
+    {
+      "epoch": 0.11279229711141678,
+      "grad_norm": 0.7766104340553284,
+      "learning_rate": 0.0009822345875271884,
+      "loss": 2.5348,
+      "step": 246
+    },
+    {
+      "epoch": 0.11325080238422743,
+      "grad_norm": 0.8359735012054443,
+      "learning_rate": 0.0009820378394501481,
+      "loss": 2.5871,
+      "step": 247
+    },
+    {
+      "epoch": 0.11370930765703806,
+      "grad_norm": 0.8868134617805481,
+      "learning_rate": 0.0009818400278178636,
+      "loss": 2.5248,
+      "step": 248
+    },
+    {
+      "epoch": 0.11416781292984869,
+      "grad_norm": 0.8946746587753296,
+      "learning_rate": 0.0009816411530667814,
+      "loss": 2.5014,
+      "step": 249
+    },
+    {
+      "epoch": 0.11462631820265932,
+      "grad_norm": 1.170770287513733,
+      "learning_rate": 0.000981441215635693,
+      "loss": 2.6139,
+      "step": 250
+    },
+    {
+      "epoch": 0.11508482347546997,
+      "grad_norm": 0.9527420401573181,
+      "learning_rate": 0.0009812402159657353,
+      "loss": 2.6185,
+      "step": 251
+    },
+    {
+      "epoch": 0.1155433287482806,
+      "grad_norm": 1.0065629482269287,
+      "learning_rate": 0.000981038154500388,
+      "loss": 2.6347,
+      "step": 252
+    },
+    {
+      "epoch": 0.11600183402109124,
+      "grad_norm": 0.9124364256858826,
+      "learning_rate": 0.0009808350316854746,
+      "loss": 2.6089,
+      "step": 253
+    },
+    {
+      "epoch": 0.11646033929390188,
+      "grad_norm": 0.9670346975326538,
+      "learning_rate": 0.0009806308479691594,
+      "loss": 2.6001,
+      "step": 254
+    },
+    {
+      "epoch": 0.11691884456671252,
+      "grad_norm": 0.7629925608634949,
+      "learning_rate": 0.0009804256038019482,
+      "loss": 2.4568,
+      "step": 255
+    },
+    {
+      "epoch": 0.11737734983952315,
+      "grad_norm": 1.0129269361495972,
+      "learning_rate": 0.0009802192996366857,
+      "loss": 2.5016,
+      "step": 256
+    },
+    {
+      "epoch": 0.1178358551123338,
+      "grad_norm": 0.7923449277877808,
+      "learning_rate": 0.0009800119359285563,
+      "loss": 2.5361,
+      "step": 257
+    },
+    {
+      "epoch": 0.11829436038514443,
+      "grad_norm": 0.8806050419807434,
+      "learning_rate": 0.0009798035131350813,
+      "loss": 2.5363,
+      "step": 258
+    },
+    {
+      "epoch": 0.11875286565795506,
+      "grad_norm": 0.8976274728775024,
+      "learning_rate": 0.0009795940317161194,
+      "loss": 2.4702,
+      "step": 259
+    },
+    {
+      "epoch": 0.1192113709307657,
+      "grad_norm": 0.8944005966186523,
+      "learning_rate": 0.0009793834921338646,
+      "loss": 2.5735,
+      "step": 260
+    },
+    {
+      "epoch": 0.11966987620357634,
+      "grad_norm": 0.8382654190063477,
+      "learning_rate": 0.0009791718948528457,
+      "loss": 2.615,
+      "step": 261
+    },
+    {
+      "epoch": 0.12012838147638698,
+      "grad_norm": 0.9910136461257935,
+      "learning_rate": 0.0009789592403399252,
+      "loss": 2.4924,
+      "step": 262
+    },
+    {
+      "epoch": 0.12058688674919761,
+      "grad_norm": 0.9050196409225464,
+      "learning_rate": 0.0009787455290642985,
+      "loss": 2.4805,
+      "step": 263
+    },
+    {
+      "epoch": 0.12104539202200826,
+      "grad_norm": 0.9110697507858276,
+      "learning_rate": 0.000978530761497492,
+      "loss": 2.5905,
+      "step": 264
+    },
+    {
+      "epoch": 0.12150389729481889,
+      "grad_norm": 0.8477030396461487,
+      "learning_rate": 0.0009783149381133633,
+      "loss": 2.4675,
+      "step": 265
+    },
+    {
+      "epoch": 0.12196240256762952,
+      "grad_norm": 0.970863938331604,
+      "learning_rate": 0.0009780980593880992,
+      "loss": 2.4856,
+      "step": 266
+    },
+    {
+      "epoch": 0.12242090784044017,
+      "grad_norm": 0.8225898742675781,
+      "learning_rate": 0.0009778801258002153,
+      "loss": 2.4673,
+      "step": 267
+    },
+    {
+      "epoch": 0.1228794131132508,
+      "grad_norm": 0.8891953825950623,
+      "learning_rate": 0.000977661137830554,
+      "loss": 2.4867,
+      "step": 268
+    },
+    {
+      "epoch": 0.12333791838606144,
+      "grad_norm": 0.9434213638305664,
+      "learning_rate": 0.0009774410959622845,
+      "loss": 2.4622,
+      "step": 269
+    },
+    {
+      "epoch": 0.12379642365887207,
+      "grad_norm": 0.8750183582305908,
+      "learning_rate": 0.000977220000680901,
+      "loss": 2.4131,
+      "step": 270
+    },
+    {
+      "epoch": 0.12425492893168272,
+      "grad_norm": 1.0006518363952637,
+      "learning_rate": 0.000976997852474223,
+      "loss": 2.4212,
+      "step": 271
+    },
+    {
+      "epoch": 0.12471343420449335,
+      "grad_norm": 0.8963906764984131,
+      "learning_rate": 0.0009767746518323914,
+      "loss": 2.4446,
+      "step": 272
+    },
+    {
+      "epoch": 0.12517193947730398,
+      "grad_norm": 0.8173921704292297,
+      "learning_rate": 0.0009765503992478704,
+      "loss": 2.3904,
+      "step": 273
+    },
+    {
+      "epoch": 0.12563044475011462,
+      "grad_norm": 0.8512886166572571,
+      "learning_rate": 0.0009763250952154449,
+      "loss": 2.4277,
+      "step": 274
+    },
+    {
+      "epoch": 0.12608895002292525,
+      "grad_norm": 0.985991895198822,
+      "learning_rate": 0.0009760987402322195,
+      "loss": 2.4364,
+      "step": 275
+    },
+    {
+      "epoch": 0.1265474552957359,
+      "grad_norm": 0.8566312789916992,
+      "learning_rate": 0.0009758713347976178,
+      "loss": 2.4763,
+      "step": 276
+    },
+    {
+      "epoch": 0.12700596056854654,
+      "grad_norm": 0.8851800560951233,
+      "learning_rate": 0.000975642879413381,
+      "loss": 2.4456,
+      "step": 277
+    },
+    {
+      "epoch": 0.12746446584135718,
+      "grad_norm": 0.9072892665863037,
+      "learning_rate": 0.0009754133745835665,
+      "loss": 2.4277,
+      "step": 278
+    },
+    {
+      "epoch": 0.1279229711141678,
+      "grad_norm": 0.8912001252174377,
+      "learning_rate": 0.0009751828208145482,
+      "loss": 2.4118,
+      "step": 279
+    },
+    {
+      "epoch": 0.12838147638697844,
+      "grad_norm": 0.9056828618049622,
+      "learning_rate": 0.0009749512186150131,
+      "loss": 2.4865,
+      "step": 280
+    },
+    {
+      "epoch": 0.12883998165978908,
+      "grad_norm": 0.959968090057373,
+      "learning_rate": 0.0009747185684959625,
+      "loss": 2.46,
+      "step": 281
+    },
+    {
+      "epoch": 0.12929848693259974,
+      "grad_norm": 0.9535790085792542,
+      "learning_rate": 0.000974484870970709,
+      "loss": 2.4158,
+      "step": 282
+    },
+    {
+      "epoch": 0.12975699220541037,
+      "grad_norm": 0.9544013738632202,
+      "learning_rate": 0.0009742501265548767,
+      "loss": 2.3775,
+      "step": 283
+    },
+    {
+      "epoch": 0.130215497478221,
+      "grad_norm": 0.9536207914352417,
+      "learning_rate": 0.0009740143357663993,
+      "loss": 2.4581,
+      "step": 284
+    },
+    {
+      "epoch": 0.13067400275103164,
+      "grad_norm": 0.9788037538528442,
+      "learning_rate": 0.000973777499125519,
+      "loss": 2.3837,
+      "step": 285
+    },
+    {
+      "epoch": 0.13113250802384227,
+      "grad_norm": 0.9597623348236084,
+      "learning_rate": 0.0009735396171547859,
+      "loss": 2.3823,
+      "step": 286
+    },
+    {
+      "epoch": 0.1315910132966529,
+      "grad_norm": 0.9844682216644287,
+      "learning_rate": 0.0009733006903790564,
+      "loss": 2.4309,
+      "step": 287
+    },
+    {
+      "epoch": 0.13204951856946354,
+      "grad_norm": 1.0646336078643799,
+      "learning_rate": 0.0009730607193254922,
+      "loss": 2.4429,
+      "step": 288
+    },
+    {
+      "epoch": 0.1325080238422742,
+      "grad_norm": 0.8027645349502563,
+      "learning_rate": 0.0009728197045235585,
+      "loss": 2.4426,
+      "step": 289
+    },
+    {
+      "epoch": 0.13296652911508483,
+      "grad_norm": 1.0418767929077148,
+      "learning_rate": 0.0009725776465050242,
+      "loss": 2.4449,
+      "step": 290
+    },
+    {
+      "epoch": 0.13342503438789546,
+      "grad_norm": 0.8396481275558472,
+      "learning_rate": 0.0009723345458039594,
+      "loss": 2.3656,
+      "step": 291
+    },
+    {
+      "epoch": 0.1338835396607061,
+      "grad_norm": 0.9335566163063049,
+      "learning_rate": 0.000972090402956735,
+      "loss": 2.3848,
+      "step": 292
+    },
+    {
+      "epoch": 0.13434204493351673,
+      "grad_norm": 0.8852063417434692,
+      "learning_rate": 0.0009718452185020212,
+      "loss": 2.3408,
+      "step": 293
+    },
+    {
+      "epoch": 0.13480055020632736,
+      "grad_norm": 1.0468389987945557,
+      "learning_rate": 0.0009715989929807862,
+      "loss": 2.3568,
+      "step": 294
+    },
+    {
+      "epoch": 0.13525905547913802,
+      "grad_norm": 0.9045504927635193,
+      "learning_rate": 0.0009713517269362955,
+      "loss": 2.3999,
+      "step": 295
+    },
+    {
+      "epoch": 0.13571756075194866,
+      "grad_norm": 0.8884844779968262,
+      "learning_rate": 0.0009711034209141101,
+      "loss": 2.4009,
+      "step": 296
+    },
+    {
+      "epoch": 0.1361760660247593,
+      "grad_norm": 0.9446419477462769,
+      "learning_rate": 0.0009708540754620856,
+      "loss": 2.4272,
+      "step": 297
+    },
+    {
+      "epoch": 0.13663457129756992,
+      "grad_norm": 0.9066890478134155,
+      "learning_rate": 0.0009706036911303713,
+      "loss": 2.4169,
+      "step": 298
+    },
+    {
+      "epoch": 0.13709307657038056,
+      "grad_norm": 1.0050914287567139,
+      "learning_rate": 0.0009703522684714083,
+      "loss": 2.3385,
+      "step": 299
+    },
+    {
+      "epoch": 0.1375515818431912,
+      "grad_norm": 0.8947672843933105,
+      "learning_rate": 0.0009700998080399286,
+      "loss": 2.3818,
+      "step": 300
+    },
+    {
+      "epoch": 0.13801008711600182,
+      "grad_norm": 0.8593084216117859,
+      "learning_rate": 0.0009698463103929542,
+      "loss": 2.405,
+      "step": 301
+    },
+    {
+      "epoch": 0.13846859238881248,
+      "grad_norm": 0.9497812986373901,
+      "learning_rate": 0.0009695917760897954,
+      "loss": 2.3784,
+      "step": 302
+    },
+    {
+      "epoch": 0.13892709766162312,
+      "grad_norm": 0.8225311636924744,
+      "learning_rate": 0.0009693362056920501,
+      "loss": 2.3898,
+      "step": 303
+    },
+    {
+      "epoch": 0.13938560293443375,
+      "grad_norm": 0.9201606512069702,
+      "learning_rate": 0.0009690795997636015,
+      "loss": 2.3531,
+      "step": 304
+    },
+    {
+      "epoch": 0.13984410820724438,
+      "grad_norm": 0.857223391532898,
+      "learning_rate": 0.0009688219588706179,
+      "loss": 2.3035,
+      "step": 305
+    },
+    {
+      "epoch": 0.14030261348005502,
+      "grad_norm": 0.8690143823623657,
+      "learning_rate": 0.0009685632835815518,
+      "loss": 2.3221,
+      "step": 306
+    },
+    {
+      "epoch": 0.14076111875286565,
+      "grad_norm": 0.9017691016197205,
+      "learning_rate": 0.0009683035744671367,
+      "loss": 2.3562,
+      "step": 307
+    },
+    {
+      "epoch": 0.14121962402567628,
+      "grad_norm": 1.0731825828552246,
+      "learning_rate": 0.0009680428321003883,
+      "loss": 2.3651,
+      "step": 308
+    },
+    {
+      "epoch": 0.14167812929848694,
+      "grad_norm": 1.0229724645614624,
+      "learning_rate": 0.000967781057056601,
+      "loss": 2.3845,
+      "step": 309
+    },
+    {
+      "epoch": 0.14213663457129758,
+      "grad_norm": 0.8972911238670349,
+      "learning_rate": 0.0009675182499133485,
+      "loss": 2.3853,
+      "step": 310
+    },
+    {
+      "epoch": 0.1425951398441082,
+      "grad_norm": 0.7953419089317322,
+      "learning_rate": 0.0009672544112504813,
+      "loss": 2.3055,
+      "step": 311
+    },
+    {
+      "epoch": 0.14305364511691884,
+      "grad_norm": 0.8953444361686707,
+      "learning_rate": 0.0009669895416501257,
+      "loss": 2.3354,
+      "step": 312
+    },
+    {
+      "epoch": 0.14351215038972948,
+      "grad_norm": 0.8807662725448608,
+      "learning_rate": 0.0009667236416966833,
+      "loss": 2.3924,
+      "step": 313
+    },
+    {
+      "epoch": 0.1439706556625401,
+      "grad_norm": 0.9119178652763367,
+      "learning_rate": 0.0009664567119768281,
+      "loss": 2.3884,
+      "step": 314
+    },
+    {
+      "epoch": 0.14442916093535077,
+      "grad_norm": 0.847438395023346,
+      "learning_rate": 0.0009661887530795067,
+      "loss": 2.2963,
+      "step": 315
+    },
+    {
+      "epoch": 0.1448876662081614,
+      "grad_norm": 0.8341944217681885,
+      "learning_rate": 0.0009659197655959365,
+      "loss": 2.345,
+      "step": 316
+    },
+    {
+      "epoch": 0.14534617148097204,
+      "grad_norm": 0.9736082553863525,
+      "learning_rate": 0.000965649750119604,
+      "loss": 2.4004,
+      "step": 317
+    },
+    {
+      "epoch": 0.14580467675378267,
+      "grad_norm": 0.9091759920120239,
+      "learning_rate": 0.0009653787072462643,
+      "loss": 2.3017,
+      "step": 318
+    },
+    {
+      "epoch": 0.1462631820265933,
+      "grad_norm": 0.9500412344932556,
+      "learning_rate": 0.0009651066375739388,
+      "loss": 2.3532,
+      "step": 319
+    },
+    {
+      "epoch": 0.14672168729940394,
+      "grad_norm": 0.8779663443565369,
+      "learning_rate": 0.000964833541702915,
+      "loss": 2.2891,
+      "step": 320
+    },
+    {
+      "epoch": 0.14718019257221457,
+      "grad_norm": 0.8973827362060547,
+      "learning_rate": 0.0009645594202357438,
+      "loss": 2.3269,
+      "step": 321
+    },
+    {
+      "epoch": 0.14763869784502523,
+      "grad_norm": 0.8694807291030884,
+      "learning_rate": 0.0009642842737772397,
+      "loss": 2.3124,
+      "step": 322
+    },
+    {
+      "epoch": 0.14809720311783586,
+      "grad_norm": 0.7520583271980286,
+      "learning_rate": 0.0009640081029344782,
+      "loss": 2.3258,
+      "step": 323
+    },
+    {
+      "epoch": 0.1485557083906465,
+      "grad_norm": 0.8970927000045776,
+      "learning_rate": 0.0009637309083167956,
+      "loss": 2.3243,
+      "step": 324
+    },
+    {
+      "epoch": 0.14901421366345713,
+      "grad_norm": 0.8834981918334961,
+      "learning_rate": 0.0009634526905357859,
+      "loss": 2.3511,
+      "step": 325
+    },
+    {
+      "epoch": 0.14947271893626776,
+      "grad_norm": 0.9334129095077515,
+      "learning_rate": 0.000963173450205302,
+      "loss": 2.3733,
+      "step": 326
+    },
+    {
+      "epoch": 0.1499312242090784,
+      "grad_norm": 0.7775444388389587,
+      "learning_rate": 0.0009628931879414517,
+      "loss": 2.2571,
+      "step": 327
+    },
+    {
+      "epoch": 0.15038972948188903,
+      "grad_norm": 0.9488975405693054,
+      "learning_rate": 0.0009626119043625983,
+      "loss": 2.3712,
+      "step": 328
+    },
+    {
+      "epoch": 0.1508482347546997,
+      "grad_norm": 0.7745993137359619,
+      "learning_rate": 0.0009623296000893582,
+      "loss": 2.2772,
+      "step": 329
+    },
+    {
+      "epoch": 0.15130674002751032,
+      "grad_norm": 0.9074013829231262,
+      "learning_rate": 0.0009620462757446,
+      "loss": 2.2865,
+      "step": 330
+    },
+    {
+      "epoch": 0.15176524530032096,
+      "grad_norm": 0.8559544086456299,
+      "learning_rate": 0.0009617619319534428,
+      "loss": 2.3847,
+      "step": 331
+    },
+    {
+      "epoch": 0.1522237505731316,
+      "grad_norm": 0.8856382966041565,
+      "learning_rate": 0.000961476569343255,
+      "loss": 2.3063,
+      "step": 332
+    },
+    {
+      "epoch": 0.15268225584594222,
+      "grad_norm": 0.8337352871894836,
+      "learning_rate": 0.0009611901885436529,
+      "loss": 2.3063,
+      "step": 333
+    },
+    {
+      "epoch": 0.15314076111875286,
+      "grad_norm": 0.7430352568626404,
+      "learning_rate": 0.0009609027901864996,
+      "loss": 2.2167,
+      "step": 334
+    },
+    {
+      "epoch": 0.15359926639156352,
+      "grad_norm": 0.8177013397216797,
+      "learning_rate": 0.0009606143749059029,
+      "loss": 2.2799,
+      "step": 335
+    },
+    {
+      "epoch": 0.15405777166437415,
+      "grad_norm": 0.743145227432251,
+      "learning_rate": 0.0009603249433382144,
+      "loss": 2.313,
+      "step": 336
+    },
+    {
+      "epoch": 0.15451627693718478,
+      "grad_norm": 0.8646201491355896,
+      "learning_rate": 0.0009600344961220282,
+      "loss": 2.3258,
+      "step": 337
+    },
+    {
+      "epoch": 0.15497478220999542,
+      "grad_norm": 0.8705004453659058,
+      "learning_rate": 0.0009597430338981791,
+      "loss": 2.2443,
+      "step": 338
+    },
+    {
+      "epoch": 0.15543328748280605,
+      "grad_norm": 0.8706442713737488,
+      "learning_rate": 0.0009594505573097414,
+      "loss": 2.247,
+      "step": 339
+    },
+    {
+      "epoch": 0.15589179275561668,
+      "grad_norm": 0.7996535897254944,
+      "learning_rate": 0.0009591570670020277,
+      "loss": 2.3581,
+      "step": 340
+    },
+    {
+      "epoch": 0.15635029802842731,
+      "grad_norm": 0.8731130361557007,
+      "learning_rate": 0.0009588625636225871,
+      "loss": 2.2238,
+      "step": 341
+    },
+    {
+      "epoch": 0.15680880330123798,
+      "grad_norm": 0.8134293556213379,
+      "learning_rate": 0.0009585670478212036,
+      "loss": 2.1063,
+      "step": 342
+    },
+    {
+      "epoch": 0.1572673085740486,
+      "grad_norm": 0.7917614579200745,
+      "learning_rate": 0.0009582705202498956,
+      "loss": 2.2579,
+      "step": 343
+    },
+    {
+      "epoch": 0.15772581384685924,
+      "grad_norm": 0.8131535649299622,
+      "learning_rate": 0.0009579729815629133,
+      "loss": 2.3116,
+      "step": 344
+    },
+    {
+      "epoch": 0.15818431911966988,
+      "grad_norm": 0.8998814821243286,
+      "learning_rate": 0.0009576744324167379,
+      "loss": 2.3828,
+      "step": 345
+    },
+    {
+      "epoch": 0.1586428243924805,
+      "grad_norm": 0.7245515584945679,
+      "learning_rate": 0.0009573748734700804,
+      "loss": 2.2207,
+      "step": 346
+    },
+    {
+      "epoch": 0.15910132966529114,
+      "grad_norm": 0.7819699645042419,
+      "learning_rate": 0.0009570743053838796,
+      "loss": 2.3543,
+      "step": 347
+    },
+    {
+      "epoch": 0.15955983493810177,
+      "grad_norm": 0.8506840467453003,
+      "learning_rate": 0.0009567727288213005,
+      "loss": 2.283,
+      "step": 348
+    },
+    {
+      "epoch": 0.16001834021091244,
+      "grad_norm": 0.7845314741134644,
+      "learning_rate": 0.0009564701444477337,
+      "loss": 2.2916,
+      "step": 349
+    },
+    {
+      "epoch": 0.16047684548372307,
+      "grad_norm": 0.7496329545974731,
+      "learning_rate": 0.000956166552930793,
+      "loss": 2.2612,
+      "step": 350
+    },
+    {
+      "epoch": 0.1609353507565337,
+      "grad_norm": 0.9572161436080933,
+      "learning_rate": 0.0009558619549403147,
+      "loss": 2.4045,
+      "step": 351
+    },
+    {
+      "epoch": 0.16139385602934433,
+      "grad_norm": 0.7328593134880066,
+      "learning_rate": 0.0009555563511483555,
+      "loss": 2.2511,
+      "step": 352
+    },
+    {
+      "epoch": 0.16185236130215497,
+      "grad_norm": 0.7712945342063904,
+      "learning_rate": 0.0009552497422291912,
+      "loss": 2.2424,
+      "step": 353
+    },
+    {
+      "epoch": 0.1623108665749656,
+      "grad_norm": 0.8544804453849792,
+      "learning_rate": 0.0009549421288593157,
+      "loss": 2.2977,
+      "step": 354
+    },
+    {
+      "epoch": 0.16276937184777626,
+      "grad_norm": 0.7544561624526978,
+      "learning_rate": 0.0009546335117174385,
+      "loss": 2.3136,
+      "step": 355
+    },
+    {
+      "epoch": 0.1632278771205869,
+      "grad_norm": 0.7910962700843811,
+      "learning_rate": 0.0009543238914844843,
+      "loss": 2.3124,
+      "step": 356
+    },
+    {
+      "epoch": 0.16368638239339753,
+      "grad_norm": 0.8093752264976501,
+      "learning_rate": 0.0009540132688435907,
+      "loss": 2.2597,
+      "step": 357
+    },
+    {
+      "epoch": 0.16414488766620816,
+      "grad_norm": 0.8010714650154114,
+      "learning_rate": 0.0009537016444801074,
+      "loss": 2.1901,
+      "step": 358
+    },
+    {
+      "epoch": 0.1646033929390188,
+      "grad_norm": 0.8825679421424866,
+      "learning_rate": 0.0009533890190815935,
+      "loss": 2.3792,
+      "step": 359
+    },
+    {
+      "epoch": 0.16506189821182943,
+      "grad_norm": 0.7667873501777649,
+      "learning_rate": 0.0009530753933378173,
+      "loss": 2.2578,
+      "step": 360
+    },
+    {
+      "epoch": 0.16552040348464006,
+      "grad_norm": 0.7517592310905457,
+      "learning_rate": 0.0009527607679407545,
+      "loss": 2.2751,
+      "step": 361
+    },
+    {
+      "epoch": 0.16597890875745072,
+      "grad_norm": 0.6640016436576843,
+      "learning_rate": 0.0009524451435845857,
+      "loss": 2.2589,
+      "step": 362
+    },
+    {
+      "epoch": 0.16643741403026135,
+      "grad_norm": 0.7178091406822205,
+      "learning_rate": 0.0009521285209656963,
+      "loss": 2.2737,
+      "step": 363
+    },
+    {
+      "epoch": 0.166895919303072,
+      "grad_norm": 0.7296462655067444,
+      "learning_rate": 0.0009518109007826734,
+      "loss": 2.1672,
+      "step": 364
+    },
+    {
+      "epoch": 0.16735442457588262,
+      "grad_norm": 0.8042681217193604,
+      "learning_rate": 0.0009514922837363059,
+      "loss": 2.2681,
+      "step": 365
+    },
+    {
+      "epoch": 0.16781292984869325,
+      "grad_norm": 0.756880521774292,
+      "learning_rate": 0.0009511726705295817,
+      "loss": 2.2504,
+      "step": 366
+    },
+    {
+      "epoch": 0.1682714351215039,
+      "grad_norm": 0.8112584352493286,
+      "learning_rate": 0.000950852061867687,
+      "loss": 2.28,
+      "step": 367
+    },
+    {
+      "epoch": 0.16872994039431455,
+      "grad_norm": 0.8013713359832764,
+      "learning_rate": 0.0009505304584580038,
+      "loss": 2.22,
+      "step": 368
+    },
+    {
+      "epoch": 0.16918844566712518,
+      "grad_norm": 0.778104305267334,
+      "learning_rate": 0.0009502078610101092,
+      "loss": 2.3591,
+      "step": 369
+    },
+    {
+      "epoch": 0.16964695093993581,
+      "grad_norm": 0.6536003351211548,
+      "learning_rate": 0.0009498842702357736,
+      "loss": 2.3031,
+      "step": 370
+    },
+    {
+      "epoch": 0.17010545621274645,
+      "grad_norm": 0.7399254441261292,
+      "learning_rate": 0.0009495596868489587,
+      "loss": 2.2491,
+      "step": 371
+    },
+    {
+      "epoch": 0.17056396148555708,
+      "grad_norm": 0.7947290539741516,
+      "learning_rate": 0.0009492341115658167,
+      "loss": 2.2877,
+      "step": 372
+    },
+    {
+      "epoch": 0.1710224667583677,
+      "grad_norm": 0.7105071544647217,
+      "learning_rate": 0.0009489075451046879,
+      "loss": 2.2605,
+      "step": 373
+    },
+    {
+      "epoch": 0.17148097203117835,
+      "grad_norm": 0.7353443503379822,
+      "learning_rate": 0.0009485799881861,
+      "loss": 2.183,
+      "step": 374
+    },
+    {
+      "epoch": 0.171939477303989,
+      "grad_norm": 0.7985981702804565,
+      "learning_rate": 0.0009482514415327654,
+      "loss": 2.2962,
+      "step": 375
+    },
+    {
+      "epoch": 0.17239798257679964,
+      "grad_norm": 0.7362819910049438,
+      "learning_rate": 0.000947921905869581,
+      "loss": 2.219,
+      "step": 376
+    },
+    {
+      "epoch": 0.17285648784961027,
+      "grad_norm": 0.769498884677887,
+      "learning_rate": 0.0009475913819236248,
+      "loss": 2.3432,
+      "step": 377
+    },
+    {
+      "epoch": 0.1733149931224209,
+      "grad_norm": 0.7538216710090637,
+      "learning_rate": 0.0009472598704241561,
+      "loss": 2.2919,
+      "step": 378
+    },
+    {
+      "epoch": 0.17377349839523154,
+      "grad_norm": 0.7968547940254211,
+      "learning_rate": 0.0009469273721026131,
+      "loss": 2.206,
+      "step": 379
+    },
+    {
+      "epoch": 0.17423200366804217,
+      "grad_norm": 0.7574167251586914,
+      "learning_rate": 0.0009465938876926111,
+      "loss": 2.2004,
+      "step": 380
+    },
+    {
+      "epoch": 0.1746905089408528,
+      "grad_norm": 0.7480878829956055,
+      "learning_rate": 0.0009462594179299406,
+      "loss": 2.345,
+      "step": 381
+    },
+    {
+      "epoch": 0.17514901421366347,
+      "grad_norm": 0.7985141277313232,
+      "learning_rate": 0.0009459239635525672,
+      "loss": 2.2559,
+      "step": 382
+    },
+    {
+      "epoch": 0.1756075194864741,
+      "grad_norm": 0.8155332803726196,
+      "learning_rate": 0.0009455875253006281,
+      "loss": 2.2384,
+      "step": 383
+    },
+    {
+      "epoch": 0.17606602475928473,
+      "grad_norm": 0.7257934808731079,
+      "learning_rate": 0.0009452501039164315,
+      "loss": 2.3794,
+      "step": 384
+    },
+    {
+      "epoch": 0.17652453003209537,
+      "grad_norm": 0.6381332874298096,
+      "learning_rate": 0.0009449117001444549,
+      "loss": 2.176,
+      "step": 385
+    },
+    {
+      "epoch": 0.176983035304906,
+      "grad_norm": 0.7162456512451172,
+      "learning_rate": 0.0009445723147313433,
+      "loss": 2.3075,
+      "step": 386
+    },
+    {
+      "epoch": 0.17744154057771663,
+      "grad_norm": 0.702751636505127,
+      "learning_rate": 0.0009442319484259074,
+      "loss": 2.1807,
+      "step": 387
+    },
+    {
+      "epoch": 0.1779000458505273,
+      "grad_norm": 0.6507847309112549,
+      "learning_rate": 0.0009438906019791222,
+      "loss": 2.3123,
+      "step": 388
+    },
+    {
+      "epoch": 0.17835855112333793,
+      "grad_norm": 0.7726557850837708,
+      "learning_rate": 0.0009435482761441251,
+      "loss": 2.3295,
+      "step": 389
+    },
+    {
+      "epoch": 0.17881705639614856,
+      "grad_norm": 0.8222095370292664,
+      "learning_rate": 0.000943204971676215,
+      "loss": 2.2905,
+      "step": 390
+    },
+    {
+      "epoch": 0.1792755616689592,
+      "grad_norm": 0.7096737027168274,
+      "learning_rate": 0.0009428606893328493,
+      "loss": 2.2274,
+      "step": 391
+    },
+    {
+      "epoch": 0.17973406694176983,
+      "grad_norm": 0.7012096643447876,
+      "learning_rate": 0.0009425154298736432,
+      "loss": 2.2492,
+      "step": 392
+    },
+    {
+      "epoch": 0.18019257221458046,
+      "grad_norm": 0.6866661906242371,
+      "learning_rate": 0.0009421691940603678,
+      "loss": 2.2075,
+      "step": 393
+    },
+    {
+      "epoch": 0.1806510774873911,
+      "grad_norm": 0.686640202999115,
+      "learning_rate": 0.0009418219826569488,
+      "loss": 2.3404,
+      "step": 394
+    },
+    {
+      "epoch": 0.18110958276020175,
+      "grad_norm": 0.7104650139808655,
+      "learning_rate": 0.0009414737964294635,
+      "loss": 2.1282,
+      "step": 395
+    },
+    {
+      "epoch": 0.1815680880330124,
+      "grad_norm": 0.73348468542099,
+      "learning_rate": 0.000941124636146141,
+      "loss": 2.2584,
+      "step": 396
+    },
+    {
+      "epoch": 0.18202659330582302,
+      "grad_norm": 0.7885403037071228,
+      "learning_rate": 0.0009407745025773589,
+      "loss": 2.1586,
+      "step": 397
+    },
+    {
+      "epoch": 0.18248509857863365,
+      "grad_norm": 0.7409605383872986,
+      "learning_rate": 0.0009404233964956423,
+      "loss": 2.1115,
+      "step": 398
+    },
+    {
+      "epoch": 0.1829436038514443,
+      "grad_norm": 0.7365392446517944,
+      "learning_rate": 0.0009400713186756625,
+      "loss": 2.2815,
+      "step": 399
+    },
+    {
+      "epoch": 0.18340210912425492,
+      "grad_norm": 0.7650797963142395,
+      "learning_rate": 0.0009397182698942342,
+      "loss": 2.1912,
+      "step": 400
+    },
+    {
+      "epoch": 0.18386061439706555,
+      "grad_norm": 0.6853351593017578,
+      "learning_rate": 0.0009393642509303149,
+      "loss": 2.2331,
+      "step": 401
+    },
+    {
+      "epoch": 0.1843191196698762,
+      "grad_norm": 0.6919862627983093,
+      "learning_rate": 0.0009390092625650023,
+      "loss": 2.211,
+      "step": 402
+    },
+    {
+      "epoch": 0.18477762494268685,
+      "grad_norm": 0.645986795425415,
+      "learning_rate": 0.0009386533055815332,
+      "loss": 2.204,
+      "step": 403
+    },
+    {
+      "epoch": 0.18523613021549748,
+      "grad_norm": 0.6640390753746033,
+      "learning_rate": 0.0009382963807652813,
+      "loss": 2.2248,
+      "step": 404
+    },
+    {
+      "epoch": 0.1856946354883081,
+      "grad_norm": 0.6698750853538513,
+      "learning_rate": 0.000937938488903756,
+      "loss": 2.2047,
+      "step": 405
+    },
+    {
+      "epoch": 0.18615314076111875,
+      "grad_norm": 0.7267535924911499,
+      "learning_rate": 0.0009375796307866003,
+      "loss": 2.2251,
+      "step": 406
+    },
+    {
+      "epoch": 0.18661164603392938,
+      "grad_norm": 0.6915977597236633,
+      "learning_rate": 0.0009372198072055888,
+      "loss": 2.2478,
+      "step": 407
+    },
+    {
+      "epoch": 0.18707015130674004,
+      "grad_norm": 0.7325314879417419,
+      "learning_rate": 0.0009368590189546268,
+      "loss": 2.1617,
+      "step": 408
+    },
+    {
+      "epoch": 0.18752865657955067,
+      "grad_norm": 0.736012876033783,
+      "learning_rate": 0.0009364972668297474,
+      "loss": 2.2568,
+      "step": 409
+    },
+    {
+      "epoch": 0.1879871618523613,
+      "grad_norm": 0.6793667674064636,
+      "learning_rate": 0.0009361345516291111,
+      "loss": 2.1792,
+      "step": 410
+    },
+    {
+      "epoch": 0.18844566712517194,
+      "grad_norm": 0.7399201989173889,
+      "learning_rate": 0.0009357708741530024,
+      "loss": 2.2521,
+      "step": 411
+    },
+    {
+      "epoch": 0.18890417239798257,
+      "grad_norm": 0.7048085927963257,
+      "learning_rate": 0.00093540623520383,
+      "loss": 2.2037,
+      "step": 412
+    },
+    {
+      "epoch": 0.1893626776707932,
+      "grad_norm": 0.7561178803443909,
+      "learning_rate": 0.000935040635586123,
+      "loss": 2.2401,
+      "step": 413
+    },
+    {
+      "epoch": 0.18982118294360384,
+      "grad_norm": 0.6761924028396606,
+      "learning_rate": 0.0009346740761065305,
+      "loss": 2.1902,
+      "step": 414
+    },
+    {
+      "epoch": 0.1902796882164145,
+      "grad_norm": 0.7049199342727661,
+      "learning_rate": 0.0009343065575738197,
+      "loss": 2.225,
+      "step": 415
+    },
+    {
+      "epoch": 0.19073819348922513,
+      "grad_norm": 0.6635369062423706,
+      "learning_rate": 0.0009339380807988733,
+      "loss": 2.2008,
+      "step": 416
+    },
+    {
+      "epoch": 0.19119669876203577,
+      "grad_norm": 0.6698092818260193,
+      "learning_rate": 0.0009335686465946887,
+      "loss": 2.1927,
+      "step": 417
+    },
+    {
+      "epoch": 0.1916552040348464,
+      "grad_norm": 0.7701328992843628,
+      "learning_rate": 0.0009331982557763754,
+      "loss": 2.2804,
+      "step": 418
+    },
+    {
+      "epoch": 0.19211370930765703,
+      "grad_norm": 0.6708125472068787,
+      "learning_rate": 0.0009328269091611537,
+      "loss": 2.1184,
+      "step": 419
+    },
+    {
+      "epoch": 0.19257221458046767,
+      "grad_norm": 0.6849743127822876,
+      "learning_rate": 0.0009324546075683524,
+      "loss": 2.2951,
+      "step": 420
+    },
+    {
+      "epoch": 0.19303071985327833,
+      "grad_norm": 0.7767400145530701,
+      "learning_rate": 0.0009320813518194083,
+      "loss": 2.2276,
+      "step": 421
+    },
+    {
+      "epoch": 0.19348922512608896,
+      "grad_norm": 0.7047038078308105,
+      "learning_rate": 0.0009317071427378624,
+      "loss": 2.1425,
+      "step": 422
+    },
+    {
+      "epoch": 0.1939477303988996,
+      "grad_norm": 0.7501230239868164,
+      "learning_rate": 0.0009313319811493594,
+      "loss": 2.1747,
+      "step": 423
+    },
+    {
+      "epoch": 0.19440623567171023,
+      "grad_norm": 0.7335860133171082,
+      "learning_rate": 0.000930955867881646,
+      "loss": 2.2514,
+      "step": 424
+    },
+    {
+      "epoch": 0.19486474094452086,
+      "grad_norm": 0.7214407920837402,
+      "learning_rate": 0.0009305788037645681,
+      "loss": 2.1875,
+      "step": 425
+    },
+    {
+      "epoch": 0.1953232462173315,
+      "grad_norm": 0.7064653038978577,
+      "learning_rate": 0.0009302007896300697,
+      "loss": 2.2183,
+      "step": 426
+    },
+    {
+      "epoch": 0.19578175149014213,
+      "grad_norm": 0.7004778981208801,
+      "learning_rate": 0.0009298218263121911,
+      "loss": 2.1818,
+      "step": 427
+    },
+    {
+      "epoch": 0.19624025676295279,
+      "grad_norm": 0.7053244113922119,
+      "learning_rate": 0.0009294419146470668,
+      "loss": 2.2499,
+      "step": 428
+    },
+    {
+      "epoch": 0.19669876203576342,
+      "grad_norm": 0.6384274363517761,
+      "learning_rate": 0.0009290610554729234,
+      "loss": 2.177,
+      "step": 429
+    },
+    {
+      "epoch": 0.19715726730857405,
+      "grad_norm": 0.7061018943786621,
+      "learning_rate": 0.0009286792496300784,
+      "loss": 2.2501,
+      "step": 430
+    },
+    {
+      "epoch": 0.19761577258138469,
+      "grad_norm": 0.6593593955039978,
+      "learning_rate": 0.0009282964979609379,
+      "loss": 2.2219,
+      "step": 431
+    },
+    {
+      "epoch": 0.19807427785419532,
+      "grad_norm": 0.6734021902084351,
+      "learning_rate": 0.0009279128013099947,
+      "loss": 2.2025,
+      "step": 432
+    },
+    {
+      "epoch": 0.19853278312700595,
+      "grad_norm": 0.6493218541145325,
+      "learning_rate": 0.0009275281605238268,
+      "loss": 2.2202,
+      "step": 433
+    },
+    {
+      "epoch": 0.19899128839981658,
+      "grad_norm": 0.6370416283607483,
+      "learning_rate": 0.0009271425764510953,
+      "loss": 2.2538,
+      "step": 434
+    },
+    {
+      "epoch": 0.19944979367262725,
+      "grad_norm": 0.7498826384544373,
+      "learning_rate": 0.0009267560499425423,
+      "loss": 2.133,
+      "step": 435
+    },
+    {
+      "epoch": 0.19990829894543788,
+      "grad_norm": 0.7238263487815857,
+      "learning_rate": 0.0009263685818509895,
+      "loss": 2.1442,
+      "step": 436
+    },
+    {
+      "epoch": 0.2003668042182485,
+      "grad_norm": 0.7106667757034302,
+      "learning_rate": 0.000925980173031336,
+      "loss": 2.2047,
+      "step": 437
+    },
+    {
+      "epoch": 0.20082530949105915,
+      "grad_norm": 0.7059872150421143,
+      "learning_rate": 0.0009255908243405567,
+      "loss": 2.1841,
+      "step": 438
+    },
+    {
+      "epoch": 0.20128381476386978,
+      "grad_norm": 0.7927334308624268,
+      "learning_rate": 0.0009252005366376996,
+      "loss": 2.2635,
+      "step": 439
+    },
+    {
+      "epoch": 0.2017423200366804,
+      "grad_norm": 0.8115891218185425,
+      "learning_rate": 0.0009248093107838852,
+      "loss": 2.2024,
+      "step": 440
+    },
+    {
+      "epoch": 0.20220082530949107,
+      "grad_norm": 0.7502104043960571,
+      "learning_rate": 0.0009244171476423036,
+      "loss": 2.2215,
+      "step": 441
+    },
+    {
+      "epoch": 0.2026593305823017,
+      "grad_norm": 0.7261080145835876,
+      "learning_rate": 0.0009240240480782129,
+      "loss": 2.1055,
+      "step": 442
+    },
+    {
+      "epoch": 0.20311783585511234,
+      "grad_norm": 0.677090048789978,
+      "learning_rate": 0.0009236300129589376,
+      "loss": 2.3082,
+      "step": 443
+    },
+    {
+      "epoch": 0.20357634112792297,
+      "grad_norm": 0.7030498385429382,
+      "learning_rate": 0.0009232350431538657,
+      "loss": 2.2432,
+      "step": 444
+    },
+    {
+      "epoch": 0.2040348464007336,
+      "grad_norm": 0.6846127510070801,
+      "learning_rate": 0.0009228391395344482,
+      "loss": 2.2263,
+      "step": 445
+    },
+    {
+      "epoch": 0.20449335167354424,
+      "grad_norm": 0.7009939551353455,
+      "learning_rate": 0.000922442302974196,
+      "loss": 2.2847,
+      "step": 446
+    },
+    {
+      "epoch": 0.20495185694635487,
+      "grad_norm": 0.7323954105377197,
+      "learning_rate": 0.0009220445343486785,
+      "loss": 2.2641,
+      "step": 447
+    },
+    {
+      "epoch": 0.20541036221916553,
+      "grad_norm": 0.7265954613685608,
+      "learning_rate": 0.0009216458345355217,
+      "loss": 2.2236,
+      "step": 448
+    },
+    {
+      "epoch": 0.20586886749197617,
+      "grad_norm": 0.7238561511039734,
+      "learning_rate": 0.0009212462044144061,
+      "loss": 2.2696,
+      "step": 449
+    },
+    {
+      "epoch": 0.2063273727647868,
+      "grad_norm": 0.6683318614959717,
+      "learning_rate": 0.0009208456448670648,
+      "loss": 2.1902,
+      "step": 450
+    },
+    {
+      "epoch": 0.20678587803759743,
+      "grad_norm": 0.7034298181533813,
+      "learning_rate": 0.0009204441567772816,
+      "loss": 2.2986,
+      "step": 451
+    },
+    {
+      "epoch": 0.20724438331040806,
+      "grad_norm": 0.6614246964454651,
+      "learning_rate": 0.0009200417410308888,
+      "loss": 2.1865,
+      "step": 452
+    },
+    {
+      "epoch": 0.2077028885832187,
+      "grad_norm": 0.683493435382843,
+      "learning_rate": 0.0009196383985157656,
+      "loss": 2.2148,
+      "step": 453
+    },
+    {
+      "epoch": 0.20816139385602933,
+      "grad_norm": 0.6523865461349487,
+      "learning_rate": 0.000919234130121836,
+      "loss": 2.2,
+      "step": 454
+    },
+    {
+      "epoch": 0.20861989912884,
+      "grad_norm": 0.7071042060852051,
+      "learning_rate": 0.0009188289367410672,
+      "loss": 2.1794,
+      "step": 455
+    },
+    {
+      "epoch": 0.20907840440165062,
+      "grad_norm": 0.6666319966316223,
+      "learning_rate": 0.0009184228192674666,
+      "loss": 2.1128,
+      "step": 456
+    },
+    {
+      "epoch": 0.20953690967446126,
+      "grad_norm": 0.603166401386261,
+      "learning_rate": 0.0009180157785970808,
+      "loss": 2.2594,
+      "step": 457
+    },
+    {
+      "epoch": 0.2099954149472719,
+      "grad_norm": 0.6653771996498108,
+      "learning_rate": 0.0009176078156279932,
+      "loss": 2.1639,
+      "step": 458
+    },
+    {
+      "epoch": 0.21045392022008252,
+      "grad_norm": 0.7248659133911133,
+      "learning_rate": 0.0009171989312603226,
+      "loss": 2.2196,
+      "step": 459
+    },
+    {
+      "epoch": 0.21091242549289316,
+      "grad_norm": 0.661429226398468,
+      "learning_rate": 0.0009167891263962202,
+      "loss": 2.1741,
+      "step": 460
+    },
+    {
+      "epoch": 0.21137093076570382,
+      "grad_norm": 0.7194615602493286,
+      "learning_rate": 0.0009163784019398685,
+      "loss": 2.1748,
+      "step": 461
+    },
+    {
+      "epoch": 0.21182943603851445,
+      "grad_norm": 0.6598905920982361,
+      "learning_rate": 0.0009159667587974785,
+      "loss": 2.3098,
+      "step": 462
+    },
+    {
+      "epoch": 0.21228794131132508,
+      "grad_norm": 0.6393786668777466,
+      "learning_rate": 0.0009155541978772887,
+      "loss": 2.1513,
+      "step": 463
+    },
+    {
+      "epoch": 0.21274644658413572,
+      "grad_norm": 0.705147922039032,
+      "learning_rate": 0.0009151407200895625,
+      "loss": 2.0953,
+      "step": 464
+    },
+    {
+      "epoch": 0.21320495185694635,
+      "grad_norm": 0.6931217312812805,
+      "learning_rate": 0.000914726326346586,
+      "loss": 2.1147,
+      "step": 465
+    },
+    {
+      "epoch": 0.21366345712975698,
+      "grad_norm": 0.7450565695762634,
+      "learning_rate": 0.0009143110175626661,
+      "loss": 2.2378,
+      "step": 466
+    },
+    {
+      "epoch": 0.21412196240256762,
+      "grad_norm": 0.6897702813148499,
+      "learning_rate": 0.0009138947946541291,
+      "loss": 2.2965,
+      "step": 467
+    },
+    {
+      "epoch": 0.21458046767537828,
+      "grad_norm": 0.7013731002807617,
+      "learning_rate": 0.0009134776585393181,
+      "loss": 2.2047,
+      "step": 468
+    },
+    {
+      "epoch": 0.2150389729481889,
+      "grad_norm": 0.6678890585899353,
+      "learning_rate": 0.0009130596101385906,
+      "loss": 2.2857,
+      "step": 469
+    },
+    {
+      "epoch": 0.21549747822099954,
+      "grad_norm": 0.7025542855262756,
+      "learning_rate": 0.0009126406503743174,
+      "loss": 2.2296,
+      "step": 470
+    },
+    {
+      "epoch": 0.21595598349381018,
+      "grad_norm": 0.6637585759162903,
+      "learning_rate": 0.0009122207801708802,
+      "loss": 2.1454,
+      "step": 471
+    },
+    {
+      "epoch": 0.2164144887666208,
+      "grad_norm": 0.692712664604187,
+      "learning_rate": 0.0009118000004546689,
+      "loss": 2.2381,
+      "step": 472
+    },
+    {
+      "epoch": 0.21687299403943144,
+      "grad_norm": 0.7005221843719482,
+      "learning_rate": 0.0009113783121540807,
+      "loss": 2.1293,
+      "step": 473
+    },
+    {
+      "epoch": 0.2173314993122421,
+      "grad_norm": 0.6710719466209412,
+      "learning_rate": 0.0009109557161995172,
+      "loss": 2.2512,
+      "step": 474
+    },
+    {
+      "epoch": 0.21779000458505274,
+      "grad_norm": 0.6700296998023987,
+      "learning_rate": 0.0009105322135233828,
+      "loss": 2.183,
+      "step": 475
+    },
+    {
+      "epoch": 0.21824850985786337,
+      "grad_norm": 0.6519913673400879,
+      "learning_rate": 0.0009101078050600821,
+      "loss": 2.1906,
+      "step": 476
+    },
+    {
+      "epoch": 0.218707015130674,
+      "grad_norm": 0.6798606514930725,
+      "learning_rate": 0.0009096824917460186,
+      "loss": 2.178,
+      "step": 477
+    },
+    {
+      "epoch": 0.21916552040348464,
+      "grad_norm": 0.7034133076667786,
+      "learning_rate": 0.0009092562745195921,
+      "loss": 2.2314,
+      "step": 478
+    },
+    {
+      "epoch": 0.21962402567629527,
+      "grad_norm": 0.7007942199707031,
+      "learning_rate": 0.0009088291543211967,
+      "loss": 2.1865,
+      "step": 479
+    },
+    {
+      "epoch": 0.2200825309491059,
+      "grad_norm": 0.6981944441795349,
+      "learning_rate": 0.0009084011320932188,
+      "loss": 2.1973,
+      "step": 480
+    },
+    {
+      "epoch": 0.22054103622191656,
+      "grad_norm": 0.6729029417037964,
+      "learning_rate": 0.0009079722087800352,
+      "loss": 2.135,
+      "step": 481
+    },
+    {
+      "epoch": 0.2209995414947272,
+      "grad_norm": 0.6946828961372375,
+      "learning_rate": 0.0009075423853280106,
+      "loss": 2.1884,
+      "step": 482
+    },
+    {
+      "epoch": 0.22145804676753783,
+      "grad_norm": 0.6490750908851624,
+      "learning_rate": 0.0009071116626854958,
+      "loss": 2.1551,
+      "step": 483
+    },
+    {
+      "epoch": 0.22191655204034846,
+      "grad_norm": 0.7388942837715149,
+      "learning_rate": 0.0009066800418028256,
+      "loss": 2.2131,
+      "step": 484
+    },
+    {
+      "epoch": 0.2223750573131591,
+      "grad_norm": 0.6238861680030823,
+      "learning_rate": 0.0009062475236323168,
+      "loss": 2.1949,
+      "step": 485
+    },
+    {
+      "epoch": 0.22283356258596973,
+      "grad_norm": 0.6407771110534668,
+      "learning_rate": 0.0009058141091282656,
+      "loss": 2.1729,
+      "step": 486
+    },
+    {
+      "epoch": 0.22329206785878036,
+      "grad_norm": 0.6265374422073364,
+      "learning_rate": 0.0009053797992469461,
+      "loss": 2.1808,
+      "step": 487
+    },
+    {
+      "epoch": 0.22375057313159102,
+      "grad_norm": 0.6444506645202637,
+      "learning_rate": 0.0009049445949466078,
+      "loss": 2.2131,
+      "step": 488
+    },
+    {
+      "epoch": 0.22420907840440166,
+      "grad_norm": 0.6467328667640686,
+      "learning_rate": 0.0009045084971874737,
+      "loss": 2.1742,
+      "step": 489
+    },
+    {
+      "epoch": 0.2246675836772123,
+      "grad_norm": 0.7562793493270874,
+      "learning_rate": 0.0009040715069317382,
+      "loss": 2.1178,
+      "step": 490
+    },
+    {
+      "epoch": 0.22512608895002292,
+      "grad_norm": 0.7215943336486816,
+      "learning_rate": 0.0009036336251435648,
+      "loss": 2.1933,
+      "step": 491
+    },
+    {
+      "epoch": 0.22558459422283356,
+      "grad_norm": 0.7383809089660645,
+      "learning_rate": 0.0009031948527890839,
+      "loss": 2.1481,
+      "step": 492
+    },
+    {
+      "epoch": 0.2260430994956442,
+      "grad_norm": 0.6941984295845032,
+      "learning_rate": 0.000902755190836391,
+      "loss": 2.1897,
+      "step": 493
+    },
+    {
+      "epoch": 0.22650160476845485,
+      "grad_norm": 0.7088770866394043,
+      "learning_rate": 0.0009023146402555442,
+      "loss": 2.2215,
+      "step": 494
+    },
+    {
+      "epoch": 0.22696011004126548,
+      "grad_norm": 0.7028141021728516,
+      "learning_rate": 0.0009018732020185624,
+      "loss": 2.1815,
+      "step": 495
+    },
+    {
+      "epoch": 0.22741861531407612,
+      "grad_norm": 0.6189048886299133,
+      "learning_rate": 0.0009014308770994235,
+      "loss": 2.1748,
+      "step": 496
+    },
+    {
+      "epoch": 0.22787712058688675,
+      "grad_norm": 0.6519414782524109,
+      "learning_rate": 0.0009009876664740605,
+      "loss": 2.2623,
+      "step": 497
+    },
+    {
+      "epoch": 0.22833562585969738,
+      "grad_norm": 0.6923785209655762,
+      "learning_rate": 0.0009005435711203618,
+      "loss": 2.2165,
+      "step": 498
+    },
+    {
+      "epoch": 0.22879413113250802,
+      "grad_norm": 0.6669988036155701,
+      "learning_rate": 0.000900098592018167,
+      "loss": 2.1899,
+      "step": 499
+    },
+    {
+      "epoch": 0.22925263640531865,
+      "grad_norm": 0.6576754450798035,
+      "learning_rate": 0.0008996527301492663,
+      "loss": 2.1192,
+      "step": 500
+    },
+    {
+      "epoch": 0.2297111416781293,
+      "grad_norm": 0.6740705370903015,
+      "learning_rate": 0.0008992059864973972,
+      "loss": 2.1029,
+      "step": 501
+    },
+    {
+      "epoch": 0.23016964695093994,
+      "grad_norm": 0.6726824045181274,
+      "learning_rate": 0.0008987583620482427,
+      "loss": 2.0788,
+      "step": 502
+    },
+    {
+      "epoch": 0.23062815222375058,
+      "grad_norm": 0.7758316397666931,
+      "learning_rate": 0.0008983098577894292,
+      "loss": 2.1428,
+      "step": 503
+    },
+    {
+      "epoch": 0.2310866574965612,
+      "grad_norm": 0.6605205535888672,
+      "learning_rate": 0.0008978604747105246,
+      "loss": 2.1438,
+      "step": 504
+    },
+    {
+      "epoch": 0.23154516276937184,
+      "grad_norm": 0.6482335329055786,
+      "learning_rate": 0.0008974102138030354,
+      "loss": 2.2687,
+      "step": 505
+    },
+    {
+      "epoch": 0.23200366804218248,
+      "grad_norm": 0.6886422038078308,
+      "learning_rate": 0.000896959076060405,
+      "loss": 2.1629,
+      "step": 506
+    },
+    {
+      "epoch": 0.2324621733149931,
+      "grad_norm": 0.6137783527374268,
+      "learning_rate": 0.0008965070624780116,
+      "loss": 2.1455,
+      "step": 507
+    },
+    {
+      "epoch": 0.23292067858780377,
+      "grad_norm": 0.6308463215827942,
+      "learning_rate": 0.0008960541740531658,
+      "loss": 2.1571,
+      "step": 508
+    },
+    {
+      "epoch": 0.2333791838606144,
+      "grad_norm": 0.6577746868133545,
+      "learning_rate": 0.0008956004117851083,
+      "loss": 2.1148,
+      "step": 509
+    },
+    {
+      "epoch": 0.23383768913342504,
+      "grad_norm": 0.6470906138420105,
+      "learning_rate": 0.0008951457766750079,
+      "loss": 2.2764,
+      "step": 510
+    },
+    {
+      "epoch": 0.23429619440623567,
+      "grad_norm": 0.7004483938217163,
+      "learning_rate": 0.0008946902697259593,
+      "loss": 2.1017,
+      "step": 511
+    },
+    {
+      "epoch": 0.2347546996790463,
+      "grad_norm": 0.6517966985702515,
+      "learning_rate": 0.0008942338919429805,
+      "loss": 2.2108,
+      "step": 512
+    },
+    {
+      "epoch": 0.23521320495185694,
+      "grad_norm": 0.7038133144378662,
+      "learning_rate": 0.0008937766443330113,
+      "loss": 2.177,
+      "step": 513
+    },
+    {
+      "epoch": 0.2356717102246676,
+      "grad_norm": 0.6334781646728516,
+      "learning_rate": 0.0008933185279049103,
+      "loss": 2.1848,
+      "step": 514
+    },
+    {
+      "epoch": 0.23613021549747823,
+      "grad_norm": 0.6510910987854004,
+      "learning_rate": 0.0008928595436694532,
+      "loss": 2.1438,
+      "step": 515
+    },
+    {
+      "epoch": 0.23658872077028886,
+      "grad_norm": 0.6366983652114868,
+      "learning_rate": 0.0008923996926393305,
+      "loss": 2.1857,
+      "step": 516
+    },
+    {
+      "epoch": 0.2370472260430995,
+      "grad_norm": 0.633724570274353,
+      "learning_rate": 0.0008919389758291449,
+      "loss": 2.053,
+      "step": 517
+    },
+    {
+      "epoch": 0.23750573131591013,
+      "grad_norm": 0.6967663764953613,
+      "learning_rate": 0.0008914773942554098,
+      "loss": 2.1709,
+      "step": 518
+    },
+    {
+      "epoch": 0.23796423658872076,
+      "grad_norm": 0.6758948564529419,
+      "learning_rate": 0.000891014948936546,
+      "loss": 2.1734,
+      "step": 519
+    },
+    {
+      "epoch": 0.2384227418615314,
+      "grad_norm": 0.6860511302947998,
+      "learning_rate": 0.0008905516408928804,
+      "loss": 2.1963,
+      "step": 520
+    },
+    {
+      "epoch": 0.23888124713434206,
+      "grad_norm": 0.6655228137969971,
+      "learning_rate": 0.0008900874711466434,
+      "loss": 2.2531,
+      "step": 521
+    },
+    {
+      "epoch": 0.2393397524071527,
+      "grad_norm": 0.6728112697601318,
+      "learning_rate": 0.0008896224407219666,
+      "loss": 2.1425,
+      "step": 522
+    },
+    {
+      "epoch": 0.23979825767996332,
+      "grad_norm": 0.6752417087554932,
+      "learning_rate": 0.0008891565506448804,
+      "loss": 2.1505,
+      "step": 523
+    },
+    {
+      "epoch": 0.24025676295277396,
+      "grad_norm": 0.6524655818939209,
+      "learning_rate": 0.0008886898019433122,
+      "loss": 2.1047,
+      "step": 524
+    },
+    {
+      "epoch": 0.2407152682255846,
+      "grad_norm": 0.6704076528549194,
+      "learning_rate": 0.0008882221956470836,
+      "loss": 2.2103,
+      "step": 525
+    },
+    {
+      "epoch": 0.24117377349839522,
+      "grad_norm": 0.6327609419822693,
+      "learning_rate": 0.0008877537327879086,
+      "loss": 2.1725,
+      "step": 526
+    },
+    {
+      "epoch": 0.24163227877120586,
+      "grad_norm": 0.6470811367034912,
+      "learning_rate": 0.0008872844143993908,
+      "loss": 2.0831,
+      "step": 527
+    },
+    {
+      "epoch": 0.24209078404401652,
+      "grad_norm": 0.6323941349983215,
+      "learning_rate": 0.0008868142415170218,
+      "loss": 2.0998,
+      "step": 528
+    },
+    {
+      "epoch": 0.24254928931682715,
+      "grad_norm": 0.6410027742385864,
+      "learning_rate": 0.0008863432151781781,
+      "loss": 2.1632,
+      "step": 529
+    },
+    {
+      "epoch": 0.24300779458963778,
+      "grad_norm": 0.6731299161911011,
+      "learning_rate": 0.0008858713364221195,
+      "loss": 2.1298,
+      "step": 530
+    },
+    {
+      "epoch": 0.24346629986244842,
+      "grad_norm": 0.6450693607330322,
+      "learning_rate": 0.0008853986062899868,
+      "loss": 2.1667,
+      "step": 531
+    },
+    {
+      "epoch": 0.24392480513525905,
+      "grad_norm": 0.623249351978302,
+      "learning_rate": 0.0008849250258247986,
+      "loss": 2.1403,
+      "step": 532
+    },
+    {
+      "epoch": 0.24438331040806968,
+      "grad_norm": 0.6274493336677551,
+      "learning_rate": 0.0008844505960714503,
+      "loss": 2.077,
+      "step": 533
+    },
+    {
+      "epoch": 0.24484181568088034,
+      "grad_norm": 0.734113335609436,
+      "learning_rate": 0.0008839753180767108,
+      "loss": 2.1941,
+      "step": 534
+    },
+    {
+      "epoch": 0.24530032095369098,
+      "grad_norm": 0.6688660979270935,
+      "learning_rate": 0.0008834991928892204,
+      "loss": 2.2205,
+      "step": 535
+    },
+    {
+      "epoch": 0.2457588262265016,
+      "grad_norm": 0.6205950379371643,
+      "learning_rate": 0.000883022221559489,
+      "loss": 2.1696,
+      "step": 536
+    },
+    {
+      "epoch": 0.24621733149931224,
+      "grad_norm": 0.6486721634864807,
+      "learning_rate": 0.0008825444051398934,
+      "loss": 2.1836,
+      "step": 537
+    },
+    {
+      "epoch": 0.24667583677212288,
+      "grad_norm": 0.6013586521148682,
+      "learning_rate": 0.0008820657446846745,
+      "loss": 2.2307,
+      "step": 538
+    },
+    {
+      "epoch": 0.2471343420449335,
+      "grad_norm": 0.6101259589195251,
+      "learning_rate": 0.000881586241249936,
+      "loss": 2.0852,
+      "step": 539
+    },
+    {
+      "epoch": 0.24759284731774414,
+      "grad_norm": 0.6370006203651428,
+      "learning_rate": 0.0008811058958936411,
+      "loss": 2.1742,
+      "step": 540
+    },
+    {
+      "epoch": 0.2480513525905548,
+      "grad_norm": 0.6113179326057434,
+      "learning_rate": 0.000880624709675611,
+      "loss": 2.0471,
+      "step": 541
+    },
+    {
+      "epoch": 0.24850985786336544,
+      "grad_norm": 0.6604048609733582,
+      "learning_rate": 0.000880142683657522,
+      "loss": 2.2604,
+      "step": 542
+    },
+    {
+      "epoch": 0.24896836313617607,
+      "grad_norm": 0.5907063484191895,
+      "learning_rate": 0.0008796598189029029,
+      "loss": 2.1048,
+      "step": 543
+    },
+    {
+      "epoch": 0.2494268684089867,
+      "grad_norm": 0.6541376113891602,
+      "learning_rate": 0.0008791761164771338,
+      "loss": 2.1046,
+      "step": 544
+    },
+    {
+      "epoch": 0.24988537368179733,
+      "grad_norm": 0.6534953117370605,
+      "learning_rate": 0.0008786915774474424,
+      "loss": 2.195,
+      "step": 545
+    },
+    {
+      "epoch": 0.25034387895460797,
+      "grad_norm": 0.6597118377685547,
+      "learning_rate": 0.0008782062028829027,
+      "loss": 2.2311,
+      "step": 546
+    },
+    {
+      "epoch": 0.25080238422741863,
+      "grad_norm": 0.7668104767799377,
+      "learning_rate": 0.0008777199938544318,
+      "loss": 2.1434,
+      "step": 547
+    },
+    {
+      "epoch": 0.25126088950022923,
+      "grad_norm": 0.637799084186554,
+      "learning_rate": 0.0008772329514347883,
+      "loss": 2.0892,
+      "step": 548
+    },
+    {
+      "epoch": 0.2517193947730399,
+      "grad_norm": 0.6944736838340759,
+      "learning_rate": 0.0008767450766985694,
+      "loss": 2.2466,
+      "step": 549
+    },
+    {
+      "epoch": 0.2521779000458505,
+      "grad_norm": 0.6075496077537537,
+      "learning_rate": 0.0008762563707222086,
+      "loss": 2.0948,
+      "step": 550
+    },
+    {
+      "epoch": 0.25263640531866116,
+      "grad_norm": 0.6816598176956177,
+      "learning_rate": 0.0008757668345839738,
+      "loss": 2.1699,
+      "step": 551
+    },
+    {
+      "epoch": 0.2530949105914718,
+      "grad_norm": 0.623039186000824,
+      "learning_rate": 0.0008752764693639638,
+      "loss": 2.1572,
+      "step": 552
+    },
+    {
+      "epoch": 0.2535534158642824,
+      "grad_norm": 0.6444469690322876,
+      "learning_rate": 0.0008747852761441078,
+      "loss": 2.245,
+      "step": 553
+    },
+    {
+      "epoch": 0.2540119211370931,
+      "grad_norm": 0.7217694520950317,
+      "learning_rate": 0.0008742932560081607,
+      "loss": 2.1866,
+      "step": 554
+    },
+    {
+      "epoch": 0.2544704264099037,
+      "grad_norm": 0.6031476259231567,
+      "learning_rate": 0.0008738004100417025,
+      "loss": 2.1903,
+      "step": 555
+    },
+    {
+      "epoch": 0.25492893168271435,
+      "grad_norm": 0.6348139643669128,
+      "learning_rate": 0.0008733067393321355,
+      "loss": 2.1527,
+      "step": 556
+    },
+    {
+      "epoch": 0.255387436955525,
+      "grad_norm": 0.6010840535163879,
+      "learning_rate": 0.000872812244968681,
+      "loss": 2.159,
+      "step": 557
+    },
+    {
+      "epoch": 0.2558459422283356,
+      "grad_norm": 0.6042171716690063,
+      "learning_rate": 0.0008723169280423783,
+      "loss": 2.2171,
+      "step": 558
+    },
+    {
+      "epoch": 0.2563044475011463,
+      "grad_norm": 0.599215030670166,
+      "learning_rate": 0.0008718207896460811,
+      "loss": 2.1803,
+      "step": 559
+    },
+    {
+      "epoch": 0.2567629527739569,
+      "grad_norm": 0.6373820900917053,
+      "learning_rate": 0.0008713238308744557,
+      "loss": 2.1231,
+      "step": 560
+    },
+    {
+      "epoch": 0.25722145804676755,
+      "grad_norm": 0.6668413281440735,
+      "learning_rate": 0.0008708260528239789,
+      "loss": 2.0939,
+      "step": 561
+    },
+    {
+      "epoch": 0.25767996331957815,
+      "grad_norm": 0.5991325974464417,
+      "learning_rate": 0.000870327456592934,
+      "loss": 2.1232,
+      "step": 562
+    },
+    {
+      "epoch": 0.2581384685923888,
+      "grad_norm": 0.6883406043052673,
+      "learning_rate": 0.0008698280432814107,
+      "loss": 2.1889,
+      "step": 563
+    },
+    {
+      "epoch": 0.2585969738651995,
+      "grad_norm": 0.6494483351707458,
+      "learning_rate": 0.000869327813991301,
+      "loss": 2.2119,
+      "step": 564
+    },
+    {
+      "epoch": 0.2590554791380101,
+      "grad_norm": 0.6266843676567078,
+      "learning_rate": 0.0008688267698262971,
+      "loss": 2.1661,
+      "step": 565
+    },
+    {
+      "epoch": 0.25951398441082074,
+      "grad_norm": 0.6588067412376404,
+      "learning_rate": 0.0008683249118918894,
+      "loss": 2.1624,
+      "step": 566
+    },
+    {
+      "epoch": 0.25997248968363135,
+      "grad_norm": 0.6632630228996277,
+      "learning_rate": 0.0008678222412953637,
+      "loss": 2.1507,
+      "step": 567
+    },
+    {
+      "epoch": 0.260430994956442,
+      "grad_norm": 0.6464502811431885,
+      "learning_rate": 0.0008673187591457987,
+      "loss": 2.2763,
+      "step": 568
+    },
+    {
+      "epoch": 0.2608895002292526,
+      "grad_norm": 0.7090128064155579,
+      "learning_rate": 0.0008668144665540639,
+      "loss": 2.1521,
+      "step": 569
+    },
+    {
+      "epoch": 0.2613480055020633,
+      "grad_norm": 0.6459299921989441,
+      "learning_rate": 0.0008663093646328167,
+      "loss": 2.1934,
+      "step": 570
+    },
+    {
+      "epoch": 0.26180651077487394,
+      "grad_norm": 0.6558036208152771,
+      "learning_rate": 0.0008658034544965003,
+      "loss": 2.0319,
+      "step": 571
+    },
+    {
+      "epoch": 0.26226501604768454,
+      "grad_norm": 0.6612682938575745,
+      "learning_rate": 0.0008652967372613412,
+      "loss": 2.135,
+      "step": 572
+    },
+    {
+      "epoch": 0.2627235213204952,
+      "grad_norm": 0.6100532412528992,
+      "learning_rate": 0.0008647892140453466,
+      "loss": 2.0817,
+      "step": 573
+    },
+    {
+      "epoch": 0.2631820265933058,
+      "grad_norm": 0.5992543697357178,
+      "learning_rate": 0.0008642808859683021,
+      "loss": 2.083,
+      "step": 574
+    },
+    {
+      "epoch": 0.26364053186611647,
+      "grad_norm": 0.6479603052139282,
+      "learning_rate": 0.0008637717541517689,
+      "loss": 2.1674,
+      "step": 575
+    },
+    {
+      "epoch": 0.2640990371389271,
+      "grad_norm": 0.60029137134552,
+      "learning_rate": 0.0008632618197190816,
+      "loss": 2.1479,
+      "step": 576
+    },
+    {
+      "epoch": 0.26455754241173773,
+      "grad_norm": 0.6512235403060913,
+      "learning_rate": 0.0008627510837953458,
+      "loss": 2.1848,
+      "step": 577
+    },
+    {
+      "epoch": 0.2650160476845484,
+      "grad_norm": 0.6031200885772705,
+      "learning_rate": 0.0008622395475074355,
+      "loss": 2.1955,
+      "step": 578
+    },
+    {
+      "epoch": 0.265474552957359,
+      "grad_norm": 0.6009354591369629,
+      "learning_rate": 0.0008617272119839903,
+      "loss": 2.121,
+      "step": 579
+    },
+    {
+      "epoch": 0.26593305823016966,
+      "grad_norm": 0.645689070224762,
+      "learning_rate": 0.0008612140783554136,
+      "loss": 2.2033,
+      "step": 580
+    },
+    {
+      "epoch": 0.26639156350298027,
+      "grad_norm": 0.6582375764846802,
+      "learning_rate": 0.0008607001477538696,
+      "loss": 2.114,
+      "step": 581
+    },
+    {
+      "epoch": 0.2668500687757909,
+      "grad_norm": 0.6325509548187256,
+      "learning_rate": 0.0008601854213132807,
+      "loss": 2.2147,
+      "step": 582
+    },
+    {
+      "epoch": 0.26730857404860153,
+      "grad_norm": 0.6488829255104065,
+      "learning_rate": 0.0008596699001693256,
+      "loss": 2.1177,
+      "step": 583
+    },
+    {
+      "epoch": 0.2677670793214122,
+      "grad_norm": 0.6282811164855957,
+      "learning_rate": 0.000859153585459436,
+      "loss": 2.1289,
+      "step": 584
+    },
+    {
+      "epoch": 0.26822558459422285,
+      "grad_norm": 0.5827179551124573,
+      "learning_rate": 0.0008586364783227949,
+      "loss": 2.1853,
+      "step": 585
+    },
+    {
+      "epoch": 0.26868408986703346,
+      "grad_norm": 0.5980232954025269,
+      "learning_rate": 0.0008581185799003332,
+      "loss": 2.2059,
+      "step": 586
+    },
+    {
+      "epoch": 0.2691425951398441,
+      "grad_norm": 0.599078893661499,
+      "learning_rate": 0.0008575998913347283,
+      "loss": 2.2076,
+      "step": 587
+    },
+    {
+      "epoch": 0.2696011004126547,
+      "grad_norm": 0.5875301361083984,
+      "learning_rate": 0.0008570804137704004,
+      "loss": 2.1394,
+      "step": 588
+    },
+    {
+      "epoch": 0.2700596056854654,
+      "grad_norm": 0.6111254692077637,
+      "learning_rate": 0.0008565601483535108,
+      "loss": 2.0818,
+      "step": 589
+    },
+    {
+      "epoch": 0.27051811095827605,
+      "grad_norm": 0.6123962998390198,
+      "learning_rate": 0.0008560390962319591,
+      "loss": 2.1327,
+      "step": 590
+    },
+    {
+      "epoch": 0.27097661623108665,
+      "grad_norm": 0.6192041635513306,
+      "learning_rate": 0.0008555172585553804,
+      "loss": 2.1529,
+      "step": 591
+    },
+    {
+      "epoch": 0.2714351215038973,
+      "grad_norm": 0.6323463916778564,
+      "learning_rate": 0.0008549946364751435,
+      "loss": 2.0699,
+      "step": 592
+    },
+    {
+      "epoch": 0.2718936267767079,
+      "grad_norm": 0.6426583528518677,
+      "learning_rate": 0.0008544712311443475,
+      "loss": 2.1903,
+      "step": 593
+    },
+    {
+      "epoch": 0.2723521320495186,
+      "grad_norm": 0.6252964735031128,
+      "learning_rate": 0.0008539470437178196,
+      "loss": 2.1483,
+      "step": 594
+    },
+    {
+      "epoch": 0.2728106373223292,
+      "grad_norm": 0.6107463836669922,
+      "learning_rate": 0.000853422075352113,
+      "loss": 2.1433,
+      "step": 595
+    },
+    {
+      "epoch": 0.27326914259513985,
+      "grad_norm": 0.6394081115722656,
+      "learning_rate": 0.0008528963272055035,
+      "loss": 2.1939,
+      "step": 596
+    },
+    {
+      "epoch": 0.2737276478679505,
+      "grad_norm": 0.6023075580596924,
+      "learning_rate": 0.0008523698004379877,
+      "loss": 2.1478,
+      "step": 597
+    },
+    {
+      "epoch": 0.2741861531407611,
+      "grad_norm": 0.6437680125236511,
+      "learning_rate": 0.00085184249621128,
+      "loss": 2.1045,
+      "step": 598
+    },
+    {
+      "epoch": 0.2746446584135718,
+      "grad_norm": 0.6092146635055542,
+      "learning_rate": 0.0008513144156888101,
+      "loss": 2.1343,
+      "step": 599
+    },
+    {
+      "epoch": 0.2751031636863824,
+      "grad_norm": 0.6918694972991943,
+      "learning_rate": 0.0008507855600357207,
+      "loss": 2.2696,
+      "step": 600
+    },
+    {
+      "epoch": 0.27556166895919304,
+      "grad_norm": 0.6367865204811096,
+      "learning_rate": 0.0008502559304188644,
+      "loss": 2.1064,
+      "step": 601
+    },
+    {
+      "epoch": 0.27602017423200365,
+      "grad_norm": 0.6820975542068481,
+      "learning_rate": 0.0008497255280068019,
+      "loss": 2.234,
+      "step": 602
+    },
+    {
+      "epoch": 0.2764786795048143,
+      "grad_norm": 0.6687871813774109,
+      "learning_rate": 0.0008491943539697986,
+      "loss": 2.1162,
+      "step": 603
+    },
+    {
+      "epoch": 0.27693718477762497,
+      "grad_norm": 0.6093264818191528,
+      "learning_rate": 0.0008486624094798226,
+      "loss": 2.1148,
+      "step": 604
+    },
+    {
+      "epoch": 0.2773956900504356,
+      "grad_norm": 0.6449092626571655,
+      "learning_rate": 0.0008481296957105417,
+      "loss": 2.1393,
+      "step": 605
+    },
+    {
+      "epoch": 0.27785419532324623,
+      "grad_norm": 0.624793291091919,
+      "learning_rate": 0.0008475962138373213,
+      "loss": 2.1422,
+      "step": 606
+    },
+    {
+      "epoch": 0.27831270059605684,
+      "grad_norm": 0.6685839295387268,
+      "learning_rate": 0.0008470619650372211,
+      "loss": 2.1259,
+      "step": 607
+    },
+    {
+      "epoch": 0.2787712058688675,
+      "grad_norm": 0.6278867125511169,
+      "learning_rate": 0.0008465269504889934,
+      "loss": 2.085,
+      "step": 608
+    },
+    {
+      "epoch": 0.2792297111416781,
+      "grad_norm": 0.6918105483055115,
+      "learning_rate": 0.0008459911713730799,
+      "loss": 2.1794,
+      "step": 609
+    },
+    {
+      "epoch": 0.27968821641448877,
+      "grad_norm": 0.6322029829025269,
+      "learning_rate": 0.0008454546288716089,
+      "loss": 2.1094,
+      "step": 610
+    },
+    {
+      "epoch": 0.2801467216872994,
+      "grad_norm": 0.6292043924331665,
+      "learning_rate": 0.0008449173241683935,
+      "loss": 2.1037,
+      "step": 611
+    },
+    {
+      "epoch": 0.28060522696011003,
+      "grad_norm": 0.6374667882919312,
+      "learning_rate": 0.0008443792584489281,
+      "loss": 2.1749,
+      "step": 612
+    },
+    {
+      "epoch": 0.2810637322329207,
+      "grad_norm": 0.6567649841308594,
+      "learning_rate": 0.0008438404329003863,
+      "loss": 2.1494,
+      "step": 613
+    },
+    {
+      "epoch": 0.2815222375057313,
+      "grad_norm": 0.6351649761199951,
+      "learning_rate": 0.0008433008487116183,
+      "loss": 2.1027,
+      "step": 614
+    },
+    {
+      "epoch": 0.28198074277854196,
+      "grad_norm": 0.6311390995979309,
+      "learning_rate": 0.0008427605070731481,
+      "loss": 2.0724,
+      "step": 615
+    },
+    {
+      "epoch": 0.28243924805135256,
+      "grad_norm": 0.6523634195327759,
+      "learning_rate": 0.0008422194091771708,
+      "loss": 2.1534,
+      "step": 616
+    },
+    {
+      "epoch": 0.2828977533241632,
+      "grad_norm": 0.6283302903175354,
+      "learning_rate": 0.0008416775562175503,
+      "loss": 2.1662,
+      "step": 617
+    },
+    {
+      "epoch": 0.2833562585969739,
+      "grad_norm": 0.6111835837364197,
+      "learning_rate": 0.000841134949389816,
+      "loss": 2.108,
+      "step": 618
+    },
+    {
+      "epoch": 0.2838147638697845,
+      "grad_norm": 0.606908917427063,
+      "learning_rate": 0.0008405915898911611,
+      "loss": 2.1804,
+      "step": 619
+    },
+    {
+      "epoch": 0.28427326914259515,
+      "grad_norm": 0.6267856955528259,
+      "learning_rate": 0.0008400474789204396,
+      "loss": 2.1525,
+      "step": 620
+    },
+    {
+      "epoch": 0.28473177441540576,
+      "grad_norm": 0.6156371235847473,
+      "learning_rate": 0.0008395026176781626,
+      "loss": 2.1319,
+      "step": 621
+    },
+    {
+      "epoch": 0.2851902796882164,
+      "grad_norm": 0.6254445910453796,
+      "learning_rate": 0.0008389570073664976,
+      "loss": 2.1741,
+      "step": 622
+    },
+    {
+      "epoch": 0.285648784961027,
+      "grad_norm": 0.6490429043769836,
+      "learning_rate": 0.0008384106491892642,
+      "loss": 2.172,
+      "step": 623
+    },
+    {
+      "epoch": 0.2861072902338377,
+      "grad_norm": 0.6401751041412354,
+      "learning_rate": 0.0008378635443519327,
+      "loss": 2.1169,
+      "step": 624
+    },
+    {
+      "epoch": 0.28656579550664835,
+      "grad_norm": 0.6156978011131287,
+      "learning_rate": 0.0008373156940616199,
+      "loss": 2.1428,
+      "step": 625
+    },
+    {
+      "epoch": 0.28702430077945895,
+      "grad_norm": 0.5860404372215271,
+      "learning_rate": 0.0008367670995270882,
+      "loss": 2.1915,
+      "step": 626
+    },
+    {
+      "epoch": 0.2874828060522696,
+      "grad_norm": 0.642985999584198,
+      "learning_rate": 0.0008362177619587416,
+      "loss": 2.1053,
+      "step": 627
+    },
+    {
+      "epoch": 0.2879413113250802,
+      "grad_norm": 0.5910913944244385,
+      "learning_rate": 0.0008356676825686238,
+      "loss": 2.0937,
+      "step": 628
+    },
+    {
+      "epoch": 0.2883998165978909,
+      "grad_norm": 0.6079606413841248,
+      "learning_rate": 0.0008351168625704147,
+      "loss": 2.2265,
+      "step": 629
+    },
+    {
+      "epoch": 0.28885832187070154,
+      "grad_norm": 0.6851686239242554,
+      "learning_rate": 0.0008345653031794292,
+      "loss": 2.1495,
+      "step": 630
+    },
+    {
+      "epoch": 0.28931682714351215,
+      "grad_norm": 0.6258017420768738,
+      "learning_rate": 0.0008340130056126125,
+      "loss": 2.066,
+      "step": 631
+    },
+    {
+      "epoch": 0.2897753324163228,
+      "grad_norm": 0.6179614067077637,
+      "learning_rate": 0.0008334599710885394,
+      "loss": 2.1325,
+      "step": 632
+    },
+    {
+      "epoch": 0.2902338376891334,
+      "grad_norm": 0.5925813317298889,
+      "learning_rate": 0.0008329062008274098,
+      "loss": 2.066,
+      "step": 633
+    },
+    {
+      "epoch": 0.2906923429619441,
+      "grad_norm": 0.610506534576416,
+      "learning_rate": 0.000832351696051048,
+      "loss": 2.0909,
+      "step": 634
+    },
+    {
+      "epoch": 0.2911508482347547,
+      "grad_norm": 0.6316505670547485,
+      "learning_rate": 0.000831796457982898,
+      "loss": 2.1533,
+      "step": 635
+    },
+    {
+      "epoch": 0.29160935350756534,
+      "grad_norm": 0.6356340050697327,
+      "learning_rate": 0.0008312404878480222,
+      "loss": 2.144,
+      "step": 636
+    },
+    {
+      "epoch": 0.292067858780376,
+      "grad_norm": 0.6798102259635925,
+      "learning_rate": 0.0008306837868730979,
+      "loss": 2.1578,
+      "step": 637
+    },
+    {
+      "epoch": 0.2925263640531866,
+      "grad_norm": 0.6357648968696594,
+      "learning_rate": 0.0008301263562864152,
+      "loss": 2.1308,
+      "step": 638
+    },
+    {
+      "epoch": 0.29298486932599727,
+      "grad_norm": 0.6131490468978882,
+      "learning_rate": 0.0008295681973178737,
+      "loss": 2.1818,
+      "step": 639
+    },
+    {
+      "epoch": 0.29344337459880787,
+      "grad_norm": 0.620820164680481,
+      "learning_rate": 0.0008290093111989804,
+      "loss": 2.136,
+      "step": 640
+    },
+    {
+      "epoch": 0.29390187987161853,
+      "grad_norm": 0.619581401348114,
+      "learning_rate": 0.0008284496991628465,
+      "loss": 2.1171,
+      "step": 641
+    },
+    {
+      "epoch": 0.29436038514442914,
+      "grad_norm": 0.630478024482727,
+      "learning_rate": 0.0008278893624441847,
+      "loss": 2.1538,
+      "step": 642
+    },
+    {
+      "epoch": 0.2948188904172398,
+      "grad_norm": 0.6074661612510681,
+      "learning_rate": 0.000827328302279307,
+      "loss": 2.0769,
+      "step": 643
+    },
+    {
+      "epoch": 0.29527739569005046,
+      "grad_norm": 0.6165046691894531,
+      "learning_rate": 0.0008267665199061211,
+      "loss": 2.0584,
+      "step": 644
+    },
+    {
+      "epoch": 0.29573590096286106,
+      "grad_norm": 0.6343932747840881,
+      "learning_rate": 0.0008262040165641288,
+      "loss": 2.1684,
+      "step": 645
+    },
+    {
+      "epoch": 0.2961944062356717,
+      "grad_norm": 0.5967280268669128,
+      "learning_rate": 0.0008256407934944219,
+      "loss": 2.1408,
+      "step": 646
+    },
+    {
+      "epoch": 0.29665291150848233,
+      "grad_norm": 0.6269424557685852,
+      "learning_rate": 0.0008250768519396807,
+      "loss": 2.0828,
+      "step": 647
+    },
+    {
+      "epoch": 0.297111416781293,
+      "grad_norm": 0.5752755999565125,
+      "learning_rate": 0.0008245121931441706,
+      "loss": 2.0769,
+      "step": 648
+    },
+    {
+      "epoch": 0.2975699220541036,
+      "grad_norm": 0.5707385540008545,
+      "learning_rate": 0.0008239468183537393,
+      "loss": 2.2211,
+      "step": 649
+    },
+    {
+      "epoch": 0.29802842732691426,
+      "grad_norm": 0.6472331881523132,
+      "learning_rate": 0.0008233807288158146,
+      "loss": 2.1309,
+      "step": 650
+    },
+    {
+      "epoch": 0.2984869325997249,
+      "grad_norm": 0.6083137392997742,
+      "learning_rate": 0.0008228139257794012,
+      "loss": 2.1196,
+      "step": 651
+    },
+    {
+      "epoch": 0.2989454378725355,
+      "grad_norm": 0.6148480176925659,
+      "learning_rate": 0.0008222464104950778,
+      "loss": 2.1255,
+      "step": 652
+    },
+    {
+      "epoch": 0.2994039431453462,
+      "grad_norm": 0.6241127252578735,
+      "learning_rate": 0.000821678184214995,
+      "loss": 2.1135,
+      "step": 653
+    },
+    {
+      "epoch": 0.2998624484181568,
+      "grad_norm": 0.6395455598831177,
+      "learning_rate": 0.0008211092481928716,
+      "loss": 2.0626,
+      "step": 654
+    },
+    {
+      "epoch": 0.30032095369096745,
+      "grad_norm": 0.6552785038948059,
+      "learning_rate": 0.0008205396036839927,
+      "loss": 2.1506,
+      "step": 655
+    },
+    {
+      "epoch": 0.30077945896377806,
+      "grad_norm": 0.5770664811134338,
+      "learning_rate": 0.0008199692519452069,
+      "loss": 2.2245,
+      "step": 656
+    },
+    {
+      "epoch": 0.3012379642365887,
+      "grad_norm": 0.5856772661209106,
+      "learning_rate": 0.0008193981942349224,
+      "loss": 2.093,
+      "step": 657
+    },
+    {
+      "epoch": 0.3016964695093994,
+      "grad_norm": 0.601642906665802,
+      "learning_rate": 0.0008188264318131056,
+      "loss": 2.1031,
+      "step": 658
+    },
+    {
+      "epoch": 0.30215497478221,
+      "grad_norm": 0.6132001876831055,
+      "learning_rate": 0.0008182539659412776,
+      "loss": 2.0844,
+      "step": 659
+    },
+    {
+      "epoch": 0.30261348005502064,
+      "grad_norm": 0.6752309203147888,
+      "learning_rate": 0.0008176807978825118,
+      "loss": 2.1065,
+      "step": 660
+    },
+    {
+      "epoch": 0.30307198532783125,
+      "grad_norm": 0.6332929134368896,
+      "learning_rate": 0.0008171069289014306,
+      "loss": 2.0657,
+      "step": 661
+    },
+    {
+      "epoch": 0.3035304906006419,
+      "grad_norm": 0.6279425024986267,
+      "learning_rate": 0.0008165323602642028,
+      "loss": 2.2054,
+      "step": 662
+    },
+    {
+      "epoch": 0.30398899587345257,
+      "grad_norm": 0.6496595144271851,
+      "learning_rate": 0.0008159570932385414,
+      "loss": 2.1202,
+      "step": 663
+    },
+    {
+      "epoch": 0.3044475011462632,
+      "grad_norm": 0.6082443594932556,
+      "learning_rate": 0.0008153811290936999,
+      "loss": 2.161,
+      "step": 664
+    },
+    {
+      "epoch": 0.30490600641907384,
+      "grad_norm": 0.607665479183197,
+      "learning_rate": 0.0008148044691004698,
+      "loss": 2.1482,
+      "step": 665
+    },
+    {
+      "epoch": 0.30536451169188444,
+      "grad_norm": 0.595714807510376,
+      "learning_rate": 0.0008142271145311783,
+      "loss": 2.1512,
+      "step": 666
+    },
+    {
+      "epoch": 0.3058230169646951,
+      "grad_norm": 0.639278769493103,
+      "learning_rate": 0.000813649066659685,
+      "loss": 2.1496,
+      "step": 667
+    },
+    {
+      "epoch": 0.3062815222375057,
+      "grad_norm": 0.6142746210098267,
+      "learning_rate": 0.0008130703267613787,
+      "loss": 2.1305,
+      "step": 668
+    },
+    {
+      "epoch": 0.30674002751031637,
+      "grad_norm": 0.588484525680542,
+      "learning_rate": 0.0008124908961131759,
+      "loss": 2.106,
+      "step": 669
+    },
+    {
+      "epoch": 0.30719853278312703,
+      "grad_norm": 0.5977964401245117,
+      "learning_rate": 0.0008119107759935163,
+      "loss": 2.0741,
+      "step": 670
+    },
+    {
+      "epoch": 0.30765703805593764,
+      "grad_norm": 0.5949472188949585,
+      "learning_rate": 0.0008113299676823615,
+      "loss": 2.172,
+      "step": 671
+    },
+    {
+      "epoch": 0.3081155433287483,
+      "grad_norm": 0.6105710864067078,
+      "learning_rate": 0.0008107484724611911,
+      "loss": 2.0898,
+      "step": 672
+    },
+    {
+      "epoch": 0.3085740486015589,
+      "grad_norm": 0.572673499584198,
+      "learning_rate": 0.0008101662916130006,
+      "loss": 2.157,
+      "step": 673
+    },
+    {
+      "epoch": 0.30903255387436956,
+      "grad_norm": 0.627693235874176,
+      "learning_rate": 0.0008095834264222979,
+      "loss": 2.1199,
+      "step": 674
+    },
+    {
+      "epoch": 0.30949105914718017,
+      "grad_norm": 0.5676590800285339,
+      "learning_rate": 0.0008089998781751009,
+      "loss": 2.2293,
+      "step": 675
+    },
+    {
+      "epoch": 0.30994956441999083,
+      "grad_norm": 0.6235236525535583,
+      "learning_rate": 0.0008084156481589349,
+      "loss": 2.1906,
+      "step": 676
+    },
+    {
+      "epoch": 0.3104080696928015,
+      "grad_norm": 0.6460664868354797,
+      "learning_rate": 0.0008078307376628291,
+      "loss": 2.1555,
+      "step": 677
+    },
+    {
+      "epoch": 0.3108665749656121,
+      "grad_norm": 0.6098769903182983,
+      "learning_rate": 0.0008072451479773143,
+      "loss": 2.1971,
+      "step": 678
+    },
+    {
+      "epoch": 0.31132508023842276,
+      "grad_norm": 0.5979719161987305,
+      "learning_rate": 0.0008066588803944195,
+      "loss": 2.1175,
+      "step": 679
+    },
+    {
+      "epoch": 0.31178358551123336,
+      "grad_norm": 0.6614172458648682,
+      "learning_rate": 0.0008060719362076697,
+      "loss": 2.1665,
+      "step": 680
+    },
+    {
+      "epoch": 0.312242090784044,
+      "grad_norm": 0.597113311290741,
+      "learning_rate": 0.0008054843167120826,
+      "loss": 2.0864,
+      "step": 681
+    },
+    {
+      "epoch": 0.31270059605685463,
+      "grad_norm": 0.603397786617279,
+      "learning_rate": 0.0008048960232041663,
+      "loss": 2.1708,
+      "step": 682
+    },
+    {
+      "epoch": 0.3131591013296653,
+      "grad_norm": 0.5946590900421143,
+      "learning_rate": 0.0008043070569819153,
+      "loss": 2.0859,
+      "step": 683
+    },
+    {
+      "epoch": 0.31361760660247595,
+      "grad_norm": 0.5935549139976501,
+      "learning_rate": 0.0008037174193448089,
+      "loss": 2.1565,
+      "step": 684
+    },
+    {
+      "epoch": 0.31407611187528656,
+      "grad_norm": 0.6196447014808655,
+      "learning_rate": 0.0008031271115938077,
+      "loss": 2.1225,
+      "step": 685
+    },
+    {
+      "epoch": 0.3145346171480972,
+      "grad_norm": 0.6670368909835815,
+      "learning_rate": 0.0008025361350313505,
+      "loss": 2.1374,
+      "step": 686
+    },
+    {
+      "epoch": 0.3149931224209078,
+      "grad_norm": 0.6598449349403381,
+      "learning_rate": 0.0008019444909613523,
+      "loss": 2.169,
+      "step": 687
+    },
+    {
+      "epoch": 0.3154516276937185,
+      "grad_norm": 0.632656991481781,
+      "learning_rate": 0.0008013521806892003,
+      "loss": 2.0182,
+      "step": 688
+    },
+    {
+      "epoch": 0.3159101329665291,
+      "grad_norm": 0.6501517295837402,
+      "learning_rate": 0.000800759205521752,
+      "loss": 2.1723,
+      "step": 689
+    },
+    {
+      "epoch": 0.31636863823933975,
+      "grad_norm": 0.5619286894798279,
+      "learning_rate": 0.0008001655667673318,
+      "loss": 2.2088,
+      "step": 690
+    },
+    {
+      "epoch": 0.3168271435121504,
+      "grad_norm": 0.7173649072647095,
+      "learning_rate": 0.0007995712657357279,
+      "loss": 2.1735,
+      "step": 691
+    },
+    {
+      "epoch": 0.317285648784961,
+      "grad_norm": 0.5738288760185242,
+      "learning_rate": 0.0007989763037381904,
+      "loss": 2.2677,
+      "step": 692
+    },
+    {
+      "epoch": 0.3177441540577717,
+      "grad_norm": 0.63090580701828,
+      "learning_rate": 0.0007983806820874271,
+      "loss": 2.0753,
+      "step": 693
+    },
+    {
+      "epoch": 0.3182026593305823,
+      "grad_norm": 0.7001709342002869,
+      "learning_rate": 0.0007977844020976016,
+      "loss": 2.1183,
+      "step": 694
+    },
+    {
+      "epoch": 0.31866116460339294,
+      "grad_norm": 0.6101516485214233,
+      "learning_rate": 0.00079718746508433,
+      "loss": 2.1476,
+      "step": 695
+    },
+    {
+      "epoch": 0.31911966987620355,
+      "grad_norm": 0.6354188919067383,
+      "learning_rate": 0.0007965898723646776,
+      "loss": 2.1403,
+      "step": 696
+    },
+    {
+      "epoch": 0.3195781751490142,
+      "grad_norm": 0.6127811670303345,
+      "learning_rate": 0.0007959916252571573,
+      "loss": 2.0655,
+      "step": 697
+    },
+    {
+      "epoch": 0.32003668042182487,
+      "grad_norm": 0.6308513283729553,
+      "learning_rate": 0.000795392725081725,
+      "loss": 2.0599,
+      "step": 698
+    },
+    {
+      "epoch": 0.3204951856946355,
+      "grad_norm": 0.6303369998931885,
+      "learning_rate": 0.000794793173159778,
+      "loss": 2.2099,
+      "step": 699
+    },
+    {
+      "epoch": 0.32095369096744614,
+      "grad_norm": 0.5899360775947571,
+      "learning_rate": 0.0007941929708141513,
+      "loss": 2.2161,
+      "step": 700
+    },
+    {
+      "epoch": 0.32141219624025674,
+      "grad_norm": 0.5701189041137695,
+      "learning_rate": 0.0007935921193691153,
+      "loss": 2.1334,
+      "step": 701
+    },
+    {
+      "epoch": 0.3218707015130674,
+      "grad_norm": 0.6146137118339539,
+      "learning_rate": 0.0007929906201503722,
+      "loss": 2.1365,
+      "step": 702
+    },
+    {
+      "epoch": 0.32232920678587806,
+      "grad_norm": 0.5993914604187012,
+      "learning_rate": 0.0007923884744850536,
+      "loss": 2.1826,
+      "step": 703
+    },
+    {
+      "epoch": 0.32278771205868867,
+      "grad_norm": 0.5696492791175842,
+      "learning_rate": 0.0007917856837017176,
+      "loss": 2.1007,
+      "step": 704
+    },
+    {
+      "epoch": 0.32324621733149933,
+      "grad_norm": 0.6127935647964478,
+      "learning_rate": 0.0007911822491303452,
+      "loss": 2.1552,
+      "step": 705
+    },
+    {
+      "epoch": 0.32370472260430994,
+      "grad_norm": 0.6150293350219727,
+      "learning_rate": 0.0007905781721023382,
+      "loss": 2.1184,
+      "step": 706
+    },
+    {
+      "epoch": 0.3241632278771206,
+      "grad_norm": 0.6172397136688232,
+      "learning_rate": 0.000789973453950516,
+      "loss": 2.0809,
+      "step": 707
+    },
+    {
+      "epoch": 0.3246217331499312,
+      "grad_norm": 0.6900002360343933,
+      "learning_rate": 0.000789368096009112,
+      "loss": 2.1621,
+      "step": 708
+    },
+    {
+      "epoch": 0.32508023842274186,
+      "grad_norm": 0.6212693452835083,
+      "learning_rate": 0.0007887620996137721,
+      "loss": 2.0871,
+      "step": 709
+    },
+    {
+      "epoch": 0.3255387436955525,
+      "grad_norm": 0.568787693977356,
+      "learning_rate": 0.0007881554661015497,
+      "loss": 2.0592,
+      "step": 710
+    },
+    {
+      "epoch": 0.32599724896836313,
+      "grad_norm": 0.6289831399917603,
+      "learning_rate": 0.0007875481968109051,
+      "loss": 2.1292,
+      "step": 711
+    },
+    {
+      "epoch": 0.3264557542411738,
+      "grad_norm": 0.6661450862884521,
+      "learning_rate": 0.0007869402930817007,
+      "loss": 2.178,
+      "step": 712
+    },
+    {
+      "epoch": 0.3269142595139844,
+      "grad_norm": 0.5816377401351929,
+      "learning_rate": 0.0007863317562551987,
+      "loss": 2.0182,
+      "step": 713
+    },
+    {
+      "epoch": 0.32737276478679506,
+      "grad_norm": 0.6140360832214355,
+      "learning_rate": 0.0007857225876740584,
+      "loss": 2.1234,
+      "step": 714
+    },
+    {
+      "epoch": 0.32783127005960566,
+      "grad_norm": 0.6022957563400269,
+      "learning_rate": 0.0007851127886823327,
+      "loss": 2.0342,
+      "step": 715
+    },
+    {
+      "epoch": 0.3282897753324163,
+      "grad_norm": 0.5674146413803101,
+      "learning_rate": 0.0007845023606254658,
+      "loss": 2.1001,
+      "step": 716
+    },
+    {
+      "epoch": 0.328748280605227,
+      "grad_norm": 0.6119834184646606,
+      "learning_rate": 0.0007838913048502894,
+      "loss": 2.0678,
+      "step": 717
+    },
+    {
+      "epoch": 0.3292067858780376,
+      "grad_norm": 0.6255372762680054,
+      "learning_rate": 0.0007832796227050208,
+      "loss": 2.2464,
+      "step": 718
+    },
+    {
+      "epoch": 0.32966529115084825,
+      "grad_norm": 0.6014360785484314,
+      "learning_rate": 0.0007826673155392587,
+      "loss": 2.1224,
+      "step": 719
+    },
+    {
+      "epoch": 0.33012379642365886,
+      "grad_norm": 0.6199411153793335,
+      "learning_rate": 0.000782054384703981,
+      "loss": 2.0704,
+      "step": 720
+    },
+    {
+      "epoch": 0.3305823016964695,
+      "grad_norm": 0.6176027655601501,
+      "learning_rate": 0.0007814408315515418,
+      "loss": 2.1332,
+      "step": 721
+    },
+    {
+      "epoch": 0.3310408069692801,
+      "grad_norm": 0.5926600098609924,
+      "learning_rate": 0.0007808266574356683,
+      "loss": 2.1469,
+      "step": 722
+    },
+    {
+      "epoch": 0.3314993122420908,
+      "grad_norm": 0.6355445384979248,
+      "learning_rate": 0.0007802118637114573,
+      "loss": 2.0942,
+      "step": 723
+    },
+    {
+      "epoch": 0.33195781751490144,
+      "grad_norm": 0.5896307229995728,
+      "learning_rate": 0.0007795964517353734,
+      "loss": 2.1924,
+      "step": 724
+    },
+    {
+      "epoch": 0.33241632278771205,
+      "grad_norm": 0.6319851875305176,
+      "learning_rate": 0.0007789804228652449,
+      "loss": 2.1057,
+      "step": 725
+    },
+    {
+      "epoch": 0.3328748280605227,
+      "grad_norm": 0.6082490682601929,
+      "learning_rate": 0.0007783637784602609,
+      "loss": 2.0941,
+      "step": 726
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.5976104736328125,
+      "learning_rate": 0.0007777465198809692,
+      "loss": 2.1701,
+      "step": 727
+    },
+    {
+      "epoch": 0.333791838606144,
+      "grad_norm": 0.5743541717529297,
+      "learning_rate": 0.0007771286484892722,
+      "loss": 2.1849,
+      "step": 728
+    },
+    {
+      "epoch": 0.3342503438789546,
+      "grad_norm": 0.6317039132118225,
+      "learning_rate": 0.000776510165648425,
+      "loss": 2.1668,
+      "step": 729
+    },
+    {
+      "epoch": 0.33470884915176524,
+      "grad_norm": 0.5675948262214661,
+      "learning_rate": 0.0007758910727230311,
+      "loss": 2.0577,
+      "step": 730
+    },
+    {
+      "epoch": 0.3351673544245759,
+      "grad_norm": 0.583967387676239,
+      "learning_rate": 0.0007752713710790404,
+      "loss": 2.0851,
+      "step": 731
+    },
+    {
+      "epoch": 0.3356258596973865,
+      "grad_norm": 0.6073348522186279,
+      "learning_rate": 0.0007746510620837459,
+      "loss": 2.1268,
+      "step": 732
+    },
+    {
+      "epoch": 0.33608436497019717,
+      "grad_norm": 0.5991381406784058,
+      "learning_rate": 0.0007740301471057807,
+      "loss": 2.1055,
+      "step": 733
+    },
+    {
+      "epoch": 0.3365428702430078,
+      "grad_norm": 0.5737477540969849,
+      "learning_rate": 0.0007734086275151146,
+      "loss": 2.1233,
+      "step": 734
+    },
+    {
+      "epoch": 0.33700137551581844,
+      "grad_norm": 0.5725003480911255,
+      "learning_rate": 0.0007727865046830517,
+      "loss": 2.0085,
+      "step": 735
+    },
+    {
+      "epoch": 0.3374598807886291,
+      "grad_norm": 0.5643078088760376,
+      "learning_rate": 0.0007721637799822269,
+      "loss": 2.1152,
+      "step": 736
+    },
+    {
+      "epoch": 0.3379183860614397,
+      "grad_norm": 0.6453188061714172,
+      "learning_rate": 0.0007715404547866032,
+      "loss": 2.0984,
+      "step": 737
+    },
+    {
+      "epoch": 0.33837689133425036,
+      "grad_norm": 0.6171191930770874,
+      "learning_rate": 0.0007709165304714685,
+      "loss": 2.1625,
+      "step": 738
+    },
+    {
+      "epoch": 0.33883539660706097,
+      "grad_norm": 0.5983705520629883,
+      "learning_rate": 0.0007702920084134324,
+      "loss": 2.1398,
+      "step": 739
+    },
+    {
+      "epoch": 0.33929390187987163,
+      "grad_norm": 0.5766205787658691,
+      "learning_rate": 0.0007696668899904236,
+      "loss": 2.1078,
+      "step": 740
+    },
+    {
+      "epoch": 0.33975240715268223,
+      "grad_norm": 0.6326219439506531,
+      "learning_rate": 0.0007690411765816864,
+      "loss": 2.1089,
+      "step": 741
+    },
+    {
+      "epoch": 0.3402109124254929,
+      "grad_norm": 0.5842326879501343,
+      "learning_rate": 0.0007684148695677778,
+      "loss": 2.0327,
+      "step": 742
+    },
+    {
+      "epoch": 0.34066941769830356,
+      "grad_norm": 0.5956586599349976,
+      "learning_rate": 0.000767787970330565,
+      "loss": 2.0579,
+      "step": 743
+    },
+    {
+      "epoch": 0.34112792297111416,
+      "grad_norm": 0.6140080094337463,
+      "learning_rate": 0.000767160480253221,
+      "loss": 2.175,
+      "step": 744
+    },
+    {
+      "epoch": 0.3415864282439248,
+      "grad_norm": 0.6095643639564514,
+      "learning_rate": 0.0007665324007202235,
+      "loss": 2.0467,
+      "step": 745
+    },
+    {
+      "epoch": 0.3420449335167354,
+      "grad_norm": 0.6217849850654602,
+      "learning_rate": 0.0007659037331173498,
+      "loss": 2.107,
+      "step": 746
+    },
+    {
+      "epoch": 0.3425034387895461,
+      "grad_norm": 0.5634474158287048,
+      "learning_rate": 0.0007652744788316752,
+      "loss": 2.0185,
+      "step": 747
+    },
+    {
+      "epoch": 0.3429619440623567,
+      "grad_norm": 0.5766761302947998,
+      "learning_rate": 0.0007646446392515692,
+      "loss": 2.073,
+      "step": 748
+    },
+    {
+      "epoch": 0.34342044933516735,
+      "grad_norm": 0.591926634311676,
+      "learning_rate": 0.000764014215766693,
+      "loss": 2.0977,
+      "step": 749
+    },
+    {
+      "epoch": 0.343878954607978,
+      "grad_norm": 0.5845012664794922,
+      "learning_rate": 0.0007633832097679958,
+      "loss": 2.0786,
+      "step": 750
+    },
+    {
+      "epoch": 0.3443374598807886,
+      "grad_norm": 0.5797628164291382,
+      "learning_rate": 0.0007627516226477122,
+      "loss": 2.0978,
+      "step": 751
+    },
+    {
+      "epoch": 0.3447959651535993,
+      "grad_norm": 0.5750910043716431,
+      "learning_rate": 0.0007621194557993589,
+      "loss": 2.2096,
+      "step": 752
+    },
+    {
+      "epoch": 0.3452544704264099,
+      "grad_norm": 0.580363929271698,
+      "learning_rate": 0.0007614867106177319,
+      "loss": 2.1077,
+      "step": 753
+    },
+    {
+      "epoch": 0.34571297569922055,
+      "grad_norm": 0.5970799922943115,
+      "learning_rate": 0.0007608533884989029,
+      "loss": 2.1422,
+      "step": 754
+    },
+    {
+      "epoch": 0.34617148097203115,
+      "grad_norm": 0.5925710797309875,
+      "learning_rate": 0.0007602194908402166,
+      "loss": 2.0658,
+      "step": 755
+    },
+    {
+      "epoch": 0.3466299862448418,
+      "grad_norm": 0.5956511497497559,
+      "learning_rate": 0.0007595850190402877,
+      "loss": 2.1371,
+      "step": 756
+    },
+    {
+      "epoch": 0.3470884915176525,
+      "grad_norm": 0.6007530093193054,
+      "learning_rate": 0.0007589499744989976,
+      "loss": 2.2195,
+      "step": 757
+    },
+    {
+      "epoch": 0.3475469967904631,
+      "grad_norm": 0.6203972697257996,
+      "learning_rate": 0.0007583143586174916,
+      "loss": 2.0962,
+      "step": 758
+    },
+    {
+      "epoch": 0.34800550206327374,
+      "grad_norm": 0.5765983462333679,
+      "learning_rate": 0.000757678172798175,
+      "loss": 2.1146,
+      "step": 759
+    },
+    {
+      "epoch": 0.34846400733608435,
+      "grad_norm": 0.5941845178604126,
+      "learning_rate": 0.0007570414184447112,
+      "loss": 2.0919,
+      "step": 760
+    },
+    {
+      "epoch": 0.348922512608895,
+      "grad_norm": 0.6153672933578491,
+      "learning_rate": 0.0007564040969620179,
+      "loss": 2.1151,
+      "step": 761
+    },
+    {
+      "epoch": 0.3493810178817056,
+      "grad_norm": 0.5701367259025574,
+      "learning_rate": 0.0007557662097562636,
+      "loss": 1.9783,
+      "step": 762
+    },
+    {
+      "epoch": 0.3498395231545163,
+      "grad_norm": 0.6027653813362122,
+      "learning_rate": 0.0007551277582348658,
+      "loss": 2.2076,
+      "step": 763
+    },
+    {
+      "epoch": 0.35029802842732694,
+      "grad_norm": 0.5864344835281372,
+      "learning_rate": 0.0007544887438064862,
+      "loss": 2.1074,
+      "step": 764
+    },
+    {
+      "epoch": 0.35075653370013754,
+      "grad_norm": 0.5893869996070862,
+      "learning_rate": 0.0007538491678810294,
+      "loss": 1.992,
+      "step": 765
+    },
+    {
+      "epoch": 0.3512150389729482,
+      "grad_norm": 0.5845503807067871,
+      "learning_rate": 0.0007532090318696381,
+      "loss": 2.1495,
+      "step": 766
+    },
+    {
+      "epoch": 0.3516735442457588,
+      "grad_norm": 0.584486186504364,
+      "learning_rate": 0.0007525683371846913,
+      "loss": 2.1421,
+      "step": 767
+    },
+    {
+      "epoch": 0.35213204951856947,
+      "grad_norm": 0.560272753238678,
+      "learning_rate": 0.0007519270852398001,
+      "loss": 2.1074,
+      "step": 768
+    },
+    {
+      "epoch": 0.35259055479138013,
+      "grad_norm": 0.5635749101638794,
+      "learning_rate": 0.000751285277449806,
+      "loss": 2.1309,
+      "step": 769
+    },
+    {
+      "epoch": 0.35304906006419073,
+      "grad_norm": 0.597835898399353,
+      "learning_rate": 0.0007506429152307756,
+      "loss": 2.0997,
+      "step": 770
+    },
+    {
+      "epoch": 0.3535075653370014,
+      "grad_norm": 0.5587824583053589,
+      "learning_rate": 0.00075,
+      "loss": 2.0539,
+      "step": 771
+    },
+    {
+      "epoch": 0.353966070609812,
+      "grad_norm": 0.579529345035553,
+      "learning_rate": 0.00074935653317599,
+      "loss": 2.1125,
+      "step": 772
+    },
+    {
+      "epoch": 0.35442457588262266,
+      "grad_norm": 0.5991502404212952,
+      "learning_rate": 0.000748712516178473,
+      "loss": 2.094,
+      "step": 773
+    },
+    {
+      "epoch": 0.35488308115543327,
+      "grad_norm": 0.584587037563324,
+      "learning_rate": 0.0007480679504283911,
+      "loss": 2.1467,
+      "step": 774
+    },
+    {
+      "epoch": 0.3553415864282439,
+      "grad_norm": 0.5366628766059875,
+      "learning_rate": 0.0007474228373478964,
+      "loss": 2.1144,
+      "step": 775
+    },
+    {
+      "epoch": 0.3558000917010546,
+      "grad_norm": 0.6424793004989624,
+      "learning_rate": 0.0007467771783603492,
+      "loss": 2.0506,
+      "step": 776
+    },
+    {
+      "epoch": 0.3562585969738652,
+      "grad_norm": 0.5471217036247253,
+      "learning_rate": 0.0007461309748903138,
+      "loss": 2.1225,
+      "step": 777
+    },
+    {
+      "epoch": 0.35671710224667585,
+      "grad_norm": 0.5840336680412292,
+      "learning_rate": 0.0007454842283635562,
+      "loss": 2.0885,
+      "step": 778
+    },
+    {
+      "epoch": 0.35717560751948646,
+      "grad_norm": 0.5894149541854858,
+      "learning_rate": 0.0007448369402070404,
+      "loss": 1.9738,
+      "step": 779
+    },
+    {
+      "epoch": 0.3576341127922971,
+      "grad_norm": 0.6157046556472778,
+      "learning_rate": 0.0007441891118489254,
+      "loss": 2.1066,
+      "step": 780
+    },
+    {
+      "epoch": 0.3580926180651077,
+      "grad_norm": 0.5945641994476318,
+      "learning_rate": 0.0007435407447185622,
+      "loss": 2.1095,
+      "step": 781
+    },
+    {
+      "epoch": 0.3585511233379184,
+      "grad_norm": 0.6653439998626709,
+      "learning_rate": 0.0007428918402464908,
+      "loss": 2.0629,
+      "step": 782
+    },
+    {
+      "epoch": 0.35900962861072905,
+      "grad_norm": 0.5890316963195801,
+      "learning_rate": 0.0007422423998644359,
+      "loss": 2.0177,
+      "step": 783
+    },
+    {
+      "epoch": 0.35946813388353965,
+      "grad_norm": 0.5780699849128723,
+      "learning_rate": 0.0007415924250053055,
+      "loss": 2.1005,
+      "step": 784
+    },
+    {
+      "epoch": 0.3599266391563503,
+      "grad_norm": 0.5832303166389465,
+      "learning_rate": 0.0007409419171031865,
+      "loss": 2.0905,
+      "step": 785
+    },
+    {
+      "epoch": 0.3603851444291609,
+      "grad_norm": 0.5474960803985596,
+      "learning_rate": 0.0007402908775933419,
+      "loss": 2.0194,
+      "step": 786
+    },
+    {
+      "epoch": 0.3608436497019716,
+      "grad_norm": 0.6404916644096375,
+      "learning_rate": 0.0007396393079122077,
+      "loss": 2.0087,
+      "step": 787
+    },
+    {
+      "epoch": 0.3613021549747822,
+      "grad_norm": 0.592486560344696,
+      "learning_rate": 0.0007389872094973896,
+      "loss": 2.0269,
+      "step": 788
+    },
+    {
+      "epoch": 0.36176066024759285,
+      "grad_norm": 0.62517911195755,
+      "learning_rate": 0.00073833458378766,
+      "loss": 2.079,
+      "step": 789
+    },
+    {
+      "epoch": 0.3622191655204035,
+      "grad_norm": 0.6128225326538086,
+      "learning_rate": 0.0007376814322229544,
+      "loss": 2.1872,
+      "step": 790
+    },
+    {
+      "epoch": 0.3626776707932141,
+      "grad_norm": 0.597076416015625,
+      "learning_rate": 0.0007370277562443688,
+      "loss": 2.0785,
+      "step": 791
+    },
+    {
+      "epoch": 0.3631361760660248,
+      "grad_norm": 0.6072973608970642,
+      "learning_rate": 0.0007363735572941564,
+      "loss": 2.1581,
+      "step": 792
+    },
+    {
+      "epoch": 0.3635946813388354,
+      "grad_norm": 0.581965982913971,
+      "learning_rate": 0.0007357188368157236,
+      "loss": 2.1803,
+      "step": 793
+    },
+    {
+      "epoch": 0.36405318661164604,
+      "grad_norm": 0.5558807849884033,
+      "learning_rate": 0.0007350635962536284,
+      "loss": 2.0484,
+      "step": 794
+    },
+    {
+      "epoch": 0.36451169188445665,
+      "grad_norm": 0.573723316192627,
+      "learning_rate": 0.0007344078370535756,
+      "loss": 2.0584,
+      "step": 795
+    },
+    {
+      "epoch": 0.3649701971572673,
+      "grad_norm": 0.5628274083137512,
+      "learning_rate": 0.0007337515606624148,
+      "loss": 2.0736,
+      "step": 796
+    },
+    {
+      "epoch": 0.36542870243007797,
+      "grad_norm": 0.559990644454956,
+      "learning_rate": 0.0007330947685281362,
+      "loss": 2.0439,
+      "step": 797
+    },
+    {
+      "epoch": 0.3658872077028886,
+      "grad_norm": 0.5909861326217651,
+      "learning_rate": 0.0007324374620998682,
+      "loss": 2.1119,
+      "step": 798
+    },
+    {
+      "epoch": 0.36634571297569923,
+      "grad_norm": 0.6268342137336731,
+      "learning_rate": 0.000731779642827874,
+      "loss": 2.042,
+      "step": 799
+    },
+    {
+      "epoch": 0.36680421824850984,
+      "grad_norm": 0.6002486348152161,
+      "learning_rate": 0.0007311213121635483,
+      "loss": 2.088,
+      "step": 800
+    },
+    {
+      "epoch": 0.3672627235213205,
+      "grad_norm": 0.6051323413848877,
+      "learning_rate": 0.0007304624715594139,
+      "loss": 2.0807,
+      "step": 801
+    },
+    {
+      "epoch": 0.3677212287941311,
+      "grad_norm": 0.600852370262146,
+      "learning_rate": 0.0007298031224691193,
+      "loss": 2.1036,
+      "step": 802
+    },
+    {
+      "epoch": 0.36817973406694177,
+      "grad_norm": 0.5892784595489502,
+      "learning_rate": 0.0007291432663474339,
+      "loss": 2.0846,
+      "step": 803
+    },
+    {
+      "epoch": 0.3686382393397524,
+      "grad_norm": 0.5702918171882629,
+      "learning_rate": 0.0007284829046502467,
+      "loss": 2.0867,
+      "step": 804
+    },
+    {
+      "epoch": 0.36909674461256303,
+      "grad_norm": 0.5879362225532532,
+      "learning_rate": 0.0007278220388345619,
+      "loss": 2.0985,
+      "step": 805
+    },
+    {
+      "epoch": 0.3695552498853737,
+      "grad_norm": 0.5847758650779724,
+      "learning_rate": 0.0007271606703584958,
+      "loss": 2.0882,
+      "step": 806
+    },
+    {
+      "epoch": 0.3700137551581843,
+      "grad_norm": 0.5503025650978088,
+      "learning_rate": 0.000726498800681274,
+      "loss": 2.0236,
+      "step": 807
+    },
+    {
+      "epoch": 0.37047226043099496,
+      "grad_norm": 0.6071052551269531,
+      "learning_rate": 0.0007258364312632279,
+      "loss": 2.0777,
+      "step": 808
+    },
+    {
+      "epoch": 0.3709307657038056,
+      "grad_norm": 0.5872209072113037,
+      "learning_rate": 0.0007251735635657915,
+      "loss": 2.0764,
+      "step": 809
+    },
+    {
+      "epoch": 0.3713892709766162,
+      "grad_norm": 0.595262885093689,
+      "learning_rate": 0.000724510199051498,
+      "loss": 2.0985,
+      "step": 810
+    },
+    {
+      "epoch": 0.3718477762494269,
+      "grad_norm": 0.6280025243759155,
+      "learning_rate": 0.0007238463391839769,
+      "loss": 2.0914,
+      "step": 811
+    },
+    {
+      "epoch": 0.3723062815222375,
+      "grad_norm": 0.5897212624549866,
+      "learning_rate": 0.0007231819854279508,
+      "loss": 2.1792,
+      "step": 812
+    },
+    {
+      "epoch": 0.37276478679504815,
+      "grad_norm": 0.6146259903907776,
+      "learning_rate": 0.0007225171392492316,
+      "loss": 2.0408,
+      "step": 813
+    },
+    {
+      "epoch": 0.37322329206785876,
+      "grad_norm": 0.5502868294715881,
+      "learning_rate": 0.0007218518021147182,
+      "loss": 2.1223,
+      "step": 814
+    },
+    {
+      "epoch": 0.3736817973406694,
+      "grad_norm": 0.6068534255027771,
+      "learning_rate": 0.0007211859754923923,
+      "loss": 2.0922,
+      "step": 815
+    },
+    {
+      "epoch": 0.3741403026134801,
+      "grad_norm": 0.6101025938987732,
+      "learning_rate": 0.0007205196608513158,
+      "loss": 2.0485,
+      "step": 816
+    },
+    {
+      "epoch": 0.3745988078862907,
+      "grad_norm": 0.5669591426849365,
+      "learning_rate": 0.0007198528596616272,
+      "loss": 2.1884,
+      "step": 817
+    },
+    {
+      "epoch": 0.37505731315910135,
+      "grad_norm": 0.5548491477966309,
+      "learning_rate": 0.0007191855733945387,
+      "loss": 2.0049,
+      "step": 818
+    },
+    {
+      "epoch": 0.37551581843191195,
+      "grad_norm": 0.5564523339271545,
+      "learning_rate": 0.0007185178035223327,
+      "loss": 2.0361,
+      "step": 819
+    },
+    {
+      "epoch": 0.3759743237047226,
+      "grad_norm": 0.5584644079208374,
+      "learning_rate": 0.0007178495515183583,
+      "loss": 2.1052,
+      "step": 820
+    },
+    {
+      "epoch": 0.3764328289775332,
+      "grad_norm": 0.5650953650474548,
+      "learning_rate": 0.000717180818857029,
+      "loss": 2.0698,
+      "step": 821
+    },
+    {
+      "epoch": 0.3768913342503439,
+      "grad_norm": 0.595164954662323,
+      "learning_rate": 0.0007165116070138182,
+      "loss": 1.988,
+      "step": 822
+    },
+    {
+      "epoch": 0.37734983952315454,
+      "grad_norm": 0.5892922878265381,
+      "learning_rate": 0.0007158419174652569,
+      "loss": 1.992,
+      "step": 823
+    },
+    {
+      "epoch": 0.37780834479596515,
+      "grad_norm": 0.5964775681495667,
+      "learning_rate": 0.00071517175168893,
+      "loss": 2.0424,
+      "step": 824
+    },
+    {
+      "epoch": 0.3782668500687758,
+      "grad_norm": 0.580890417098999,
+      "learning_rate": 0.0007145011111634732,
+      "loss": 2.1505,
+      "step": 825
+    },
+    {
+      "epoch": 0.3787253553415864,
+      "grad_norm": 0.6355502605438232,
+      "learning_rate": 0.0007138299973685694,
+      "loss": 2.1452,
+      "step": 826
+    },
+    {
+      "epoch": 0.3791838606143971,
+      "grad_norm": 0.6152413487434387,
+      "learning_rate": 0.0007131584117849459,
+      "loss": 2.0388,
+      "step": 827
+    },
+    {
+      "epoch": 0.3796423658872077,
+      "grad_norm": 0.5860288143157959,
+      "learning_rate": 0.0007124863558943713,
+      "loss": 2.0841,
+      "step": 828
+    },
+    {
+      "epoch": 0.38010087116001834,
+      "grad_norm": 0.5494410395622253,
+      "learning_rate": 0.0007118138311796514,
+      "loss": 2.167,
+      "step": 829
+    },
+    {
+      "epoch": 0.380559376432829,
+      "grad_norm": 0.6198868155479431,
+      "learning_rate": 0.0007111408391246262,
+      "loss": 2.0481,
+      "step": 830
+    },
+    {
+      "epoch": 0.3810178817056396,
+      "grad_norm": 0.5641661882400513,
+      "learning_rate": 0.0007104673812141675,
+      "loss": 2.0913,
+      "step": 831
+    },
+    {
+      "epoch": 0.38147638697845027,
+      "grad_norm": 0.6113286018371582,
+      "learning_rate": 0.0007097934589341745,
+      "loss": 2.1232,
+      "step": 832
+    },
+    {
+      "epoch": 0.38193489225126087,
+      "grad_norm": 0.6032219529151917,
+      "learning_rate": 0.0007091190737715711,
+      "loss": 2.1441,
+      "step": 833
+    },
+    {
+      "epoch": 0.38239339752407153,
+      "grad_norm": 0.5531289577484131,
+      "learning_rate": 0.0007084442272143026,
+      "loss": 2.1161,
+      "step": 834
+    },
+    {
+      "epoch": 0.38285190279688214,
+      "grad_norm": 0.6001880764961243,
+      "learning_rate": 0.000707768920751332,
+      "loss": 2.1568,
+      "step": 835
+    },
+    {
+      "epoch": 0.3833104080696928,
+      "grad_norm": 0.564295768737793,
+      "learning_rate": 0.0007070931558726373,
+      "loss": 2.1539,
+      "step": 836
+    },
+    {
+      "epoch": 0.38376891334250346,
+      "grad_norm": 0.5301439166069031,
+      "learning_rate": 0.0007064169340692076,
+      "loss": 2.0796,
+      "step": 837
+    },
+    {
+      "epoch": 0.38422741861531406,
+      "grad_norm": 0.602712869644165,
+      "learning_rate": 0.0007057402568330407,
+      "loss": 1.9713,
+      "step": 838
+    },
+    {
+      "epoch": 0.3846859238881247,
+      "grad_norm": 0.6009853482246399,
+      "learning_rate": 0.0007050631256571389,
+      "loss": 2.0237,
+      "step": 839
+    },
+    {
+      "epoch": 0.38514442916093533,
+      "grad_norm": 0.5657638907432556,
+      "learning_rate": 0.000704385542035506,
+      "loss": 2.0404,
+      "step": 840
+    },
+    {
+      "epoch": 0.385602934433746,
+      "grad_norm": 0.6145375967025757,
+      "learning_rate": 0.000703707507463144,
+      "loss": 2.0433,
+      "step": 841
+    },
+    {
+      "epoch": 0.38606143970655665,
+      "grad_norm": 0.6422620415687561,
+      "learning_rate": 0.0007030290234360505,
+      "loss": 2.1838,
+      "step": 842
+    },
+    {
+      "epoch": 0.38651994497936726,
+      "grad_norm": 0.542242169380188,
+      "learning_rate": 0.0007023500914512139,
+      "loss": 2.1074,
+      "step": 843
+    },
+    {
+      "epoch": 0.3869784502521779,
+      "grad_norm": 0.632345974445343,
+      "learning_rate": 0.0007016707130066116,
+      "loss": 2.0717,
+      "step": 844
+    },
+    {
+      "epoch": 0.3874369555249885,
+      "grad_norm": 0.6991784572601318,
+      "learning_rate": 0.0007009908896012055,
+      "loss": 2.0773,
+      "step": 845
+    },
+    {
+      "epoch": 0.3878954607977992,
+      "grad_norm": 0.5628587603569031,
+      "learning_rate": 0.0007003106227349399,
+      "loss": 2.1669,
+      "step": 846
+    },
+    {
+      "epoch": 0.3883539660706098,
+      "grad_norm": 0.5881508588790894,
+      "learning_rate": 0.000699629913908737,
+      "loss": 2.0963,
+      "step": 847
+    },
+    {
+      "epoch": 0.38881247134342045,
+      "grad_norm": 0.6468292474746704,
+      "learning_rate": 0.0006989487646244943,
+      "loss": 2.0978,
+      "step": 848
+    },
+    {
+      "epoch": 0.3892709766162311,
+      "grad_norm": 0.598924994468689,
+      "learning_rate": 0.0006982671763850814,
+      "loss": 2.0875,
+      "step": 849
+    },
+    {
+      "epoch": 0.3897294818890417,
+      "grad_norm": 0.6058657169342041,
+      "learning_rate": 0.0006975851506943359,
+      "loss": 2.0743,
+      "step": 850
+    },
+    {
+      "epoch": 0.3901879871618524,
+      "grad_norm": 0.5762587785720825,
+      "learning_rate": 0.0006969026890570611,
+      "loss": 2.122,
+      "step": 851
+    },
+    {
+      "epoch": 0.390646492434663,
+      "grad_norm": 0.5708557963371277,
+      "learning_rate": 0.0006962197929790216,
+      "loss": 1.9672,
+      "step": 852
+    },
+    {
+      "epoch": 0.39110499770747365,
+      "grad_norm": 0.552362322807312,
+      "learning_rate": 0.0006955364639669409,
+      "loss": 2.1306,
+      "step": 853
+    },
+    {
+      "epoch": 0.39156350298028425,
+      "grad_norm": 0.5845114588737488,
+      "learning_rate": 0.0006948527035284978,
+      "loss": 2.0687,
+      "step": 854
+    },
+    {
+      "epoch": 0.3920220082530949,
+      "grad_norm": 0.5929469466209412,
+      "learning_rate": 0.0006941685131723225,
+      "loss": 2.1481,
+      "step": 855
+    },
+    {
+      "epoch": 0.39248051352590557,
+      "grad_norm": 0.5497490763664246,
+      "learning_rate": 0.0006934838944079943,
+      "loss": 2.0966,
+      "step": 856
+    },
+    {
+      "epoch": 0.3929390187987162,
+      "grad_norm": 0.5312681198120117,
+      "learning_rate": 0.0006927988487460378,
+      "loss": 2.0693,
+      "step": 857
+    },
+    {
+      "epoch": 0.39339752407152684,
+      "grad_norm": 0.5843544006347656,
+      "learning_rate": 0.0006921133776979186,
+      "loss": 2.0723,
+      "step": 858
+    },
+    {
+      "epoch": 0.39385602934433744,
+      "grad_norm": 0.5893259644508362,
+      "learning_rate": 0.0006914274827760418,
+      "loss": 2.0495,
+      "step": 859
+    },
+    {
+      "epoch": 0.3943145346171481,
+      "grad_norm": 0.5947468280792236,
+      "learning_rate": 0.0006907411654937475,
+      "loss": 2.1046,
+      "step": 860
+    },
+    {
+      "epoch": 0.3947730398899587,
+      "grad_norm": 0.5633609890937805,
+      "learning_rate": 0.0006900544273653075,
+      "loss": 2.1804,
+      "step": 861
+    },
+    {
+      "epoch": 0.39523154516276937,
+      "grad_norm": 0.5672656893730164,
+      "learning_rate": 0.000689367269905922,
+      "loss": 2.1183,
+      "step": 862
+    },
+    {
+      "epoch": 0.39569005043558003,
+      "grad_norm": 0.5731209516525269,
+      "learning_rate": 0.0006886796946317168,
+      "loss": 2.1115,
+      "step": 863
+    },
+    {
+      "epoch": 0.39614855570839064,
+      "grad_norm": 0.5531408786773682,
+      "learning_rate": 0.0006879917030597397,
+      "loss": 2.0712,
+      "step": 864
+    },
+    {
+      "epoch": 0.3966070609812013,
+      "grad_norm": 0.6008120775222778,
+      "learning_rate": 0.0006873032967079561,
+      "loss": 2.0319,
+      "step": 865
+    },
+    {
+      "epoch": 0.3970655662540119,
+      "grad_norm": 0.5546537637710571,
+      "learning_rate": 0.0006866144770952474,
+      "loss": 2.0842,
+      "step": 866
+    },
+    {
+      "epoch": 0.39752407152682256,
+      "grad_norm": 0.5470041632652283,
+      "learning_rate": 0.0006859252457414067,
+      "loss": 2.1509,
+      "step": 867
+    },
+    {
+      "epoch": 0.39798257679963317,
+      "grad_norm": 0.5688446164131165,
+      "learning_rate": 0.0006852356041671351,
+      "loss": 2.035,
+      "step": 868
+    },
+    {
+      "epoch": 0.39844108207244383,
+      "grad_norm": 0.5843449234962463,
+      "learning_rate": 0.0006845455538940394,
+      "loss": 2.0878,
+      "step": 869
+    },
+    {
+      "epoch": 0.3988995873452545,
+      "grad_norm": 0.5773811936378479,
+      "learning_rate": 0.0006838550964446276,
+      "loss": 2.0519,
+      "step": 870
+    },
+    {
+      "epoch": 0.3993580926180651,
+      "grad_norm": 0.5943328142166138,
+      "learning_rate": 0.0006831642333423067,
+      "loss": 2.1193,
+      "step": 871
+    },
+    {
+      "epoch": 0.39981659789087576,
+      "grad_norm": 0.5568351745605469,
+      "learning_rate": 0.000682472966111378,
+      "loss": 2.0311,
+      "step": 872
+    },
+    {
+      "epoch": 0.40027510316368636,
+      "grad_norm": 0.5933377742767334,
+      "learning_rate": 0.0006817812962770348,
+      "loss": 2.0926,
+      "step": 873
+    },
+    {
+      "epoch": 0.400733608436497,
+      "grad_norm": 0.5722256302833557,
+      "learning_rate": 0.0006810892253653589,
+      "loss": 2.0834,
+      "step": 874
+    },
+    {
+      "epoch": 0.40119211370930763,
+      "grad_norm": 0.5601034760475159,
+      "learning_rate": 0.0006803967549033167,
+      "loss": 2.0706,
+      "step": 875
+    },
+    {
+      "epoch": 0.4016506189821183,
+      "grad_norm": 0.5711630582809448,
+      "learning_rate": 0.0006797038864187564,
+      "loss": 2.1273,
+      "step": 876
+    },
+    {
+      "epoch": 0.40210912425492895,
+      "grad_norm": 0.6101444363594055,
+      "learning_rate": 0.0006790106214404043,
+      "loss": 2.1408,
+      "step": 877
+    },
+    {
+      "epoch": 0.40256762952773956,
+      "grad_norm": 0.5205982327461243,
+      "learning_rate": 0.0006783169614978614,
+      "loss": 2.06,
+      "step": 878
+    },
+    {
+      "epoch": 0.4030261348005502,
+      "grad_norm": 0.6048743724822998,
+      "learning_rate": 0.0006776229081216001,
+      "loss": 2.085,
+      "step": 879
+    },
+    {
+      "epoch": 0.4034846400733608,
+      "grad_norm": 0.5251579880714417,
+      "learning_rate": 0.0006769284628429611,
+      "loss": 2.0416,
+      "step": 880
+    },
+    {
+      "epoch": 0.4039431453461715,
+      "grad_norm": 0.5934441685676575,
+      "learning_rate": 0.0006762336271941498,
+      "loss": 2.1032,
+      "step": 881
+    },
+    {
+      "epoch": 0.40440165061898214,
+      "grad_norm": 0.5310015678405762,
+      "learning_rate": 0.0006755384027082326,
+      "loss": 2.0589,
+      "step": 882
+    },
+    {
+      "epoch": 0.40486015589179275,
+      "grad_norm": 0.5640329718589783,
+      "learning_rate": 0.0006748427909191342,
+      "loss": 2.0198,
+      "step": 883
+    },
+    {
+      "epoch": 0.4053186611646034,
+      "grad_norm": 0.5701730251312256,
+      "learning_rate": 0.0006741467933616335,
+      "loss": 2.0477,
+      "step": 884
+    },
+    {
+      "epoch": 0.405777166437414,
+      "grad_norm": 0.5497170090675354,
+      "learning_rate": 0.0006734504115713604,
+      "loss": 2.0376,
+      "step": 885
+    },
+    {
+      "epoch": 0.4062356717102247,
+      "grad_norm": 0.5795982480049133,
+      "learning_rate": 0.0006727536470847932,
+      "loss": 2.0818,
+      "step": 886
+    },
+    {
+      "epoch": 0.4066941769830353,
+      "grad_norm": 0.6133224964141846,
+      "learning_rate": 0.000672056501439254,
+      "loss": 2.0647,
+      "step": 887
+    },
+    {
+      "epoch": 0.40715268225584594,
+      "grad_norm": 0.5936967134475708,
+      "learning_rate": 0.0006713589761729063,
+      "loss": 2.0435,
+      "step": 888
+    },
+    {
+      "epoch": 0.4076111875286566,
+      "grad_norm": 0.598312497138977,
+      "learning_rate": 0.0006706610728247508,
+      "loss": 2.0984,
+      "step": 889
+    },
+    {
+      "epoch": 0.4080696928014672,
+      "grad_norm": 0.569863498210907,
+      "learning_rate": 0.0006699627929346227,
+      "loss": 2.1426,
+      "step": 890
+    },
+    {
+      "epoch": 0.40852819807427787,
+      "grad_norm": 0.554601788520813,
+      "learning_rate": 0.0006692641380431879,
+      "loss": 2.0612,
+      "step": 891
+    },
+    {
+      "epoch": 0.4089867033470885,
+      "grad_norm": 0.5345975160598755,
+      "learning_rate": 0.0006685651096919393,
+      "loss": 2.0754,
+      "step": 892
+    },
+    {
+      "epoch": 0.40944520861989914,
+      "grad_norm": 0.6223799586296082,
+      "learning_rate": 0.0006678657094231944,
+      "loss": 1.8839,
+      "step": 893
+    },
+    {
+      "epoch": 0.40990371389270974,
+      "grad_norm": 0.5639466643333435,
+      "learning_rate": 0.0006671659387800909,
+      "loss": 2.2277,
+      "step": 894
+    },
+    {
+      "epoch": 0.4103622191655204,
+      "grad_norm": 0.589162290096283,
+      "learning_rate": 0.000666465799306584,
+      "loss": 2.0735,
+      "step": 895
+    },
+    {
+      "epoch": 0.41082072443833106,
+      "grad_norm": 0.5842369198799133,
+      "learning_rate": 0.0006657652925474423,
+      "loss": 2.0768,
+      "step": 896
+    },
+    {
+      "epoch": 0.41127922971114167,
+      "grad_norm": 0.6050604581832886,
+      "learning_rate": 0.000665064420048245,
+      "loss": 2.1032,
+      "step": 897
+    },
+    {
+      "epoch": 0.41173773498395233,
+      "grad_norm": 0.5710419416427612,
+      "learning_rate": 0.0006643631833553785,
+      "loss": 2.0557,
+      "step": 898
+    },
+    {
+      "epoch": 0.41219624025676294,
+      "grad_norm": 0.5624141097068787,
+      "learning_rate": 0.000663661584016032,
+      "loss": 2.0254,
+      "step": 899
+    },
+    {
+      "epoch": 0.4126547455295736,
+      "grad_norm": 0.5728394985198975,
+      "learning_rate": 0.0006629596235781957,
+      "loss": 2.1061,
+      "step": 900
+    },
+    {
+      "epoch": 0.4131132508023842,
+      "grad_norm": 0.5657985806465149,
+      "learning_rate": 0.0006622573035906556,
+      "loss": 2.0955,
+      "step": 901
+    },
+    {
+      "epoch": 0.41357175607519486,
+      "grad_norm": 0.5759608745574951,
+      "learning_rate": 0.0006615546256029921,
+      "loss": 2.0785,
+      "step": 902
+    },
+    {
+      "epoch": 0.4140302613480055,
+      "grad_norm": 0.6076114773750305,
+      "learning_rate": 0.0006608515911655743,
+      "loss": 2.2138,
+      "step": 903
+    },
+    {
+      "epoch": 0.41448876662081613,
+      "grad_norm": 0.5762312412261963,
+      "learning_rate": 0.0006601482018295591,
+      "loss": 2.1152,
+      "step": 904
+    },
+    {
+      "epoch": 0.4149472718936268,
+      "grad_norm": 0.5324718356132507,
+      "learning_rate": 0.0006594444591468851,
+      "loss": 2.1196,
+      "step": 905
+    },
+    {
+      "epoch": 0.4154057771664374,
+      "grad_norm": 0.5260560512542725,
+      "learning_rate": 0.0006587403646702713,
+      "loss": 2.1104,
+      "step": 906
+    },
+    {
+      "epoch": 0.41586428243924806,
+      "grad_norm": 0.5788759589195251,
+      "learning_rate": 0.0006580359199532126,
+      "loss": 2.0545,
+      "step": 907
+    },
+    {
+      "epoch": 0.41632278771205866,
+      "grad_norm": 0.5419743657112122,
+      "learning_rate": 0.000657331126549977,
+      "loss": 2.0173,
+      "step": 908
+    },
+    {
+      "epoch": 0.4167812929848693,
+      "grad_norm": 0.5371017456054688,
+      "learning_rate": 0.0006566259860156014,
+      "loss": 2.0449,
+      "step": 909
+    },
+    {
+      "epoch": 0.41723979825768,
+      "grad_norm": 0.5827140808105469,
+      "learning_rate": 0.0006559204999058888,
+      "loss": 2.0873,
+      "step": 910
+    },
+    {
+      "epoch": 0.4176983035304906,
+      "grad_norm": 0.5387760400772095,
+      "learning_rate": 0.0006552146697774049,
+      "loss": 2.1095,
+      "step": 911
+    },
+    {
+      "epoch": 0.41815680880330125,
+      "grad_norm": 0.5511270761489868,
+      "learning_rate": 0.0006545084971874737,
+      "loss": 2.0919,
+      "step": 912
+    },
+    {
+      "epoch": 0.41861531407611186,
+      "grad_norm": 0.5827509164810181,
+      "learning_rate": 0.0006538019836941758,
+      "loss": 2.0949,
+      "step": 913
+    },
+    {
+      "epoch": 0.4190738193489225,
+      "grad_norm": 0.6108592748641968,
+      "learning_rate": 0.0006530951308563431,
+      "loss": 2.1293,
+      "step": 914
+    },
+    {
+      "epoch": 0.4195323246217332,
+      "grad_norm": 0.5636149644851685,
+      "learning_rate": 0.0006523879402335567,
+      "loss": 2.1028,
+      "step": 915
+    },
+    {
+      "epoch": 0.4199908298945438,
+      "grad_norm": 0.5541301369667053,
+      "learning_rate": 0.0006516804133861429,
+      "loss": 2.0975,
+      "step": 916
+    },
+    {
+      "epoch": 0.42044933516735444,
+      "grad_norm": 0.5771334171295166,
+      "learning_rate": 0.0006509725518751698,
+      "loss": 2.0343,
+      "step": 917
+    },
+    {
+      "epoch": 0.42090784044016505,
+      "grad_norm": 0.6181041598320007,
+      "learning_rate": 0.0006502643572624438,
+      "loss": 2.0599,
+      "step": 918
+    },
+    {
+      "epoch": 0.4213663457129757,
+      "grad_norm": 0.5933142900466919,
+      "learning_rate": 0.0006495558311105064,
+      "loss": 2.0316,
+      "step": 919
+    },
+    {
+      "epoch": 0.4218248509857863,
+      "grad_norm": 0.569390058517456,
+      "learning_rate": 0.0006488469749826305,
+      "loss": 2.0623,
+      "step": 920
+    },
+    {
+      "epoch": 0.422283356258597,
+      "grad_norm": 0.5759417414665222,
+      "learning_rate": 0.000648137790442817,
+      "loss": 2.1245,
+      "step": 921
+    },
+    {
+      "epoch": 0.42274186153140764,
+      "grad_norm": 0.562913179397583,
+      "learning_rate": 0.0006474282790557916,
+      "loss": 2.0153,
+      "step": 922
+    },
+    {
+      "epoch": 0.42320036680421824,
+      "grad_norm": 0.5103808641433716,
+      "learning_rate": 0.000646718442387001,
+      "loss": 2.049,
+      "step": 923
+    },
+    {
+      "epoch": 0.4236588720770289,
+      "grad_norm": 0.544730007648468,
+      "learning_rate": 0.0006460082820026094,
+      "loss": 2.0338,
+      "step": 924
+    },
+    {
+      "epoch": 0.4241173773498395,
+      "grad_norm": 0.5627383589744568,
+      "learning_rate": 0.0006452977994694959,
+      "loss": 2.0054,
+      "step": 925
+    },
+    {
+      "epoch": 0.42457588262265017,
+      "grad_norm": 0.5960891246795654,
+      "learning_rate": 0.0006445869963552496,
+      "loss": 2.0365,
+      "step": 926
+    },
+    {
+      "epoch": 0.4250343878954608,
+      "grad_norm": 0.54454505443573,
+      "learning_rate": 0.0006438758742281672,
+      "loss": 2.0889,
+      "step": 927
+    },
+    {
+      "epoch": 0.42549289316827144,
+      "grad_norm": 0.5772391557693481,
+      "learning_rate": 0.0006431644346572495,
+      "loss": 2.0506,
+      "step": 928
+    },
+    {
+      "epoch": 0.4259513984410821,
+      "grad_norm": 0.5795130133628845,
+      "learning_rate": 0.0006424526792121974,
+      "loss": 1.9596,
+      "step": 929
+    },
+    {
+      "epoch": 0.4264099037138927,
+      "grad_norm": 0.5226057767868042,
+      "learning_rate": 0.0006417406094634089,
+      "loss": 2.1277,
+      "step": 930
+    },
+    {
+      "epoch": 0.42686840898670336,
+      "grad_norm": 0.5226566791534424,
+      "learning_rate": 0.0006410282269819756,
+      "loss": 2.0545,
+      "step": 931
+    },
+    {
+      "epoch": 0.42732691425951397,
+      "grad_norm": 0.5470696687698364,
+      "learning_rate": 0.0006403155333396787,
+      "loss": 2.1227,
+      "step": 932
+    },
+    {
+      "epoch": 0.42778541953232463,
+      "grad_norm": 0.5184722542762756,
+      "learning_rate": 0.0006396025301089863,
+      "loss": 2.1317,
+      "step": 933
+    },
+    {
+      "epoch": 0.42824392480513523,
+      "grad_norm": 0.5431411266326904,
+      "learning_rate": 0.0006388892188630493,
+      "loss": 2.0086,
+      "step": 934
+    },
+    {
+      "epoch": 0.4287024300779459,
+      "grad_norm": 0.5201200246810913,
+      "learning_rate": 0.0006381756011756982,
+      "loss": 2.0554,
+      "step": 935
+    },
+    {
+      "epoch": 0.42916093535075656,
+      "grad_norm": 0.5177744626998901,
+      "learning_rate": 0.0006374616786214403,
+      "loss": 2.1405,
+      "step": 936
+    },
+    {
+      "epoch": 0.42961944062356716,
+      "grad_norm": 0.5030388832092285,
+      "learning_rate": 0.0006367474527754544,
+      "loss": 2.03,
+      "step": 937
+    },
+    {
+      "epoch": 0.4300779458963778,
+      "grad_norm": 0.5276607275009155,
+      "learning_rate": 0.0006360329252135894,
+      "loss": 2.1061,
+      "step": 938
+    },
+    {
+      "epoch": 0.4305364511691884,
+      "grad_norm": 0.5480859279632568,
+      "learning_rate": 0.0006353180975123595,
+      "loss": 2.0787,
+      "step": 939
+    },
+    {
+      "epoch": 0.4309949564419991,
+      "grad_norm": 0.532832145690918,
+      "learning_rate": 0.0006346029712489413,
+      "loss": 2.1062,
+      "step": 940
+    },
+    {
+      "epoch": 0.4314534617148097,
+      "grad_norm": 0.6202355623245239,
+      "learning_rate": 0.0006338875480011698,
+      "loss": 2.1649,
+      "step": 941
+    },
+    {
+      "epoch": 0.43191196698762035,
+      "grad_norm": 0.5683257579803467,
+      "learning_rate": 0.0006331718293475357,
+      "loss": 1.9819,
+      "step": 942
+    },
+    {
+      "epoch": 0.432370472260431,
+      "grad_norm": 0.6157258749008179,
+      "learning_rate": 0.0006324558168671811,
+      "loss": 2.0392,
+      "step": 943
+    },
+    {
+      "epoch": 0.4328289775332416,
+      "grad_norm": 0.6201665997505188,
+      "learning_rate": 0.0006317395121398968,
+      "loss": 2.1208,
+      "step": 944
+    },
+    {
+      "epoch": 0.4332874828060523,
+      "grad_norm": 0.561103880405426,
+      "learning_rate": 0.0006310229167461179,
+      "loss": 2.0431,
+      "step": 945
+    },
+    {
+      "epoch": 0.4337459880788629,
+      "grad_norm": 0.6070934534072876,
+      "learning_rate": 0.0006303060322669214,
+      "loss": 1.9706,
+      "step": 946
+    },
+    {
+      "epoch": 0.43420449335167355,
+      "grad_norm": 0.6036852598190308,
+      "learning_rate": 0.0006295888602840214,
+      "loss": 2.2064,
+      "step": 947
+    },
+    {
+      "epoch": 0.4346629986244842,
+      "grad_norm": 0.5861724615097046,
+      "learning_rate": 0.0006288714023797671,
+      "loss": 2.0753,
+      "step": 948
+    },
+    {
+      "epoch": 0.4351215038972948,
+      "grad_norm": 0.5617519021034241,
+      "learning_rate": 0.000628153660137138,
+      "loss": 2.1723,
+      "step": 949
+    },
+    {
+      "epoch": 0.4355800091701055,
+      "grad_norm": 0.6019743084907532,
+      "learning_rate": 0.0006274356351397413,
+      "loss": 2.0929,
+      "step": 950
+    },
+    {
+      "epoch": 0.4360385144429161,
+      "grad_norm": 0.5480517745018005,
+      "learning_rate": 0.0006267173289718079,
+      "loss": 2.0248,
+      "step": 951
+    },
+    {
+      "epoch": 0.43649701971572674,
+      "grad_norm": 0.5686712265014648,
+      "learning_rate": 0.000625998743218189,
+      "loss": 2.0663,
+      "step": 952
+    },
+    {
+      "epoch": 0.43695552498853735,
+      "grad_norm": 0.5797551870346069,
+      "learning_rate": 0.000625279879464353,
+      "loss": 2.0287,
+      "step": 953
+    },
+    {
+      "epoch": 0.437414030261348,
+      "grad_norm": 0.6008578538894653,
+      "learning_rate": 0.000624560739296381,
+      "loss": 2.0791,
+      "step": 954
+    },
+    {
+      "epoch": 0.43787253553415867,
+      "grad_norm": 0.593769371509552,
+      "learning_rate": 0.0006238413243009648,
+      "loss": 2.041,
+      "step": 955
+    },
+    {
+      "epoch": 0.4383310408069693,
+      "grad_norm": 0.5723699331283569,
+      "learning_rate": 0.000623121636065402,
+      "loss": 2.0794,
+      "step": 956
+    },
+    {
+      "epoch": 0.43878954607977994,
+      "grad_norm": 0.6154850125312805,
+      "learning_rate": 0.0006224016761775933,
+      "loss": 2.0657,
+      "step": 957
+    },
+    {
+      "epoch": 0.43924805135259054,
+      "grad_norm": 0.6273459196090698,
+      "learning_rate": 0.0006216814462260386,
+      "loss": 2.116,
+      "step": 958
+    },
+    {
+      "epoch": 0.4397065566254012,
+      "grad_norm": 0.5462487936019897,
+      "learning_rate": 0.0006209609477998338,
+      "loss": 2.044,
+      "step": 959
+    },
+    {
+      "epoch": 0.4401650618982118,
+      "grad_norm": 0.5521886944770813,
+      "learning_rate": 0.0006202401824886674,
+      "loss": 2.0792,
+      "step": 960
+    },
+    {
+      "epoch": 0.44062356717102247,
+      "grad_norm": 0.5848297476768494,
+      "learning_rate": 0.0006195191518828162,
+      "loss": 2.0715,
+      "step": 961
+    },
+    {
+      "epoch": 0.44108207244383313,
+      "grad_norm": 0.5829888582229614,
+      "learning_rate": 0.0006187978575731427,
+      "loss": 2.0956,
+      "step": 962
+    },
+    {
+      "epoch": 0.44154057771664373,
+      "grad_norm": 0.5800769925117493,
+      "learning_rate": 0.0006180763011510911,
+      "loss": 2.0958,
+      "step": 963
+    },
+    {
+      "epoch": 0.4419990829894544,
+      "grad_norm": 0.5739432573318481,
+      "learning_rate": 0.000617354484208684,
+      "loss": 2.0146,
+      "step": 964
+    },
+    {
+      "epoch": 0.442457588262265,
+      "grad_norm": 0.5567271113395691,
+      "learning_rate": 0.0006166324083385189,
+      "loss": 2.064,
+      "step": 965
+    },
+    {
+      "epoch": 0.44291609353507566,
+      "grad_norm": 0.5500732660293579,
+      "learning_rate": 0.0006159100751337642,
+      "loss": 2.0915,
+      "step": 966
+    },
+    {
+      "epoch": 0.44337459880788627,
+      "grad_norm": 0.5444322228431702,
+      "learning_rate": 0.0006151874861881565,
+      "loss": 2.1124,
+      "step": 967
+    },
+    {
+      "epoch": 0.4438331040806969,
+      "grad_norm": 0.5586362481117249,
+      "learning_rate": 0.0006144646430959964,
+      "loss": 2.0389,
+      "step": 968
+    },
+    {
+      "epoch": 0.4442916093535076,
+      "grad_norm": 0.5322284698486328,
+      "learning_rate": 0.0006137415474521454,
+      "loss": 2.0076,
+      "step": 969
+    },
+    {
+      "epoch": 0.4447501146263182,
+      "grad_norm": 0.5360018610954285,
+      "learning_rate": 0.0006130182008520222,
+      "loss": 2.0472,
+      "step": 970
+    },
+    {
+      "epoch": 0.44520861989912885,
+      "grad_norm": 0.5501183867454529,
+      "learning_rate": 0.000612294604891599,
+      "loss": 2.0509,
+      "step": 971
+    },
+    {
+      "epoch": 0.44566712517193946,
+      "grad_norm": 0.548341691493988,
+      "learning_rate": 0.0006115707611673986,
+      "loss": 2.076,
+      "step": 972
+    },
+    {
+      "epoch": 0.4461256304447501,
+      "grad_norm": 0.5677660703659058,
+      "learning_rate": 0.0006108466712764902,
+      "loss": 2.1601,
+      "step": 973
+    },
+    {
+      "epoch": 0.4465841357175607,
+      "grad_norm": 0.5385782718658447,
+      "learning_rate": 0.0006101223368164858,
+      "loss": 2.07,
+      "step": 974
+    },
+    {
+      "epoch": 0.4470426409903714,
+      "grad_norm": 0.5334963202476501,
+      "learning_rate": 0.0006093977593855375,
+      "loss": 2.1286,
+      "step": 975
+    },
+    {
+      "epoch": 0.44750114626318205,
+      "grad_norm": 0.554218590259552,
+      "learning_rate": 0.0006086729405823335,
+      "loss": 2.031,
+      "step": 976
+    },
+    {
+      "epoch": 0.44795965153599265,
+      "grad_norm": 0.5493168234825134,
+      "learning_rate": 0.0006079478820060943,
+      "loss": 2.0243,
+      "step": 977
+    },
+    {
+      "epoch": 0.4484181568088033,
+      "grad_norm": 0.5113298296928406,
+      "learning_rate": 0.0006072225852565695,
+      "loss": 2.0952,
+      "step": 978
+    },
+    {
+      "epoch": 0.4488766620816139,
+      "grad_norm": 0.553216278553009,
+      "learning_rate": 0.0006064970519340341,
+      "loss": 2.0329,
+      "step": 979
+    },
+    {
+      "epoch": 0.4493351673544246,
+      "grad_norm": 0.5781317949295044,
+      "learning_rate": 0.0006057712836392856,
+      "loss": 2.0558,
+      "step": 980
+    },
+    {
+      "epoch": 0.4497936726272352,
+      "grad_norm": 0.551362931728363,
+      "learning_rate": 0.0006050452819736389,
+      "loss": 2.0896,
+      "step": 981
+    },
+    {
+      "epoch": 0.45025217790004585,
+      "grad_norm": 0.592470645904541,
+      "learning_rate": 0.000604319048538925,
+      "loss": 2.0612,
+      "step": 982
+    },
+    {
+      "epoch": 0.4507106831728565,
+      "grad_norm": 0.570504367351532,
+      "learning_rate": 0.0006035925849374855,
+      "loss": 2.0278,
+      "step": 983
+    },
+    {
+      "epoch": 0.4511691884456671,
+      "grad_norm": 0.5162889957427979,
+      "learning_rate": 0.0006028658927721697,
+      "loss": 2.0942,
+      "step": 984
+    },
+    {
+      "epoch": 0.4516276937184778,
+      "grad_norm": 0.5673566460609436,
+      "learning_rate": 0.0006021389736463321,
+      "loss": 2.0462,
+      "step": 985
+    },
+    {
+      "epoch": 0.4520861989912884,
+      "grad_norm": 0.609852135181427,
+      "learning_rate": 0.0006014118291638271,
+      "loss": 2.0462,
+      "step": 986
+    },
+    {
+      "epoch": 0.45254470426409904,
+      "grad_norm": 0.5628025531768799,
+      "learning_rate": 0.0006006844609290065,
+      "loss": 2.1095,
+      "step": 987
+    },
+    {
+      "epoch": 0.4530032095369097,
+      "grad_norm": 0.5801230072975159,
+      "learning_rate": 0.0005999568705467161,
+      "loss": 2.0983,
+      "step": 988
+    },
+    {
+      "epoch": 0.4534617148097203,
+      "grad_norm": 0.5281678438186646,
+      "learning_rate": 0.0005992290596222915,
+      "loss": 2.0354,
+      "step": 989
+    },
+    {
+      "epoch": 0.45392022008253097,
+      "grad_norm": 0.5476452708244324,
+      "learning_rate": 0.0005985010297615551,
+      "loss": 1.9787,
+      "step": 990
+    },
+    {
+      "epoch": 0.4543787253553416,
+      "grad_norm": 0.6053197383880615,
+      "learning_rate": 0.0005977727825708123,
+      "loss": 2.0797,
+      "step": 991
+    },
+    {
+      "epoch": 0.45483723062815223,
+      "grad_norm": 0.5676044821739197,
+      "learning_rate": 0.0005970443196568478,
+      "loss": 2.0806,
+      "step": 992
+    },
+    {
+      "epoch": 0.45529573590096284,
+      "grad_norm": 0.5678106546401978,
+      "learning_rate": 0.0005963156426269227,
+      "loss": 2.0997,
+      "step": 993
+    },
+    {
+      "epoch": 0.4557542411737735,
+      "grad_norm": 0.5809860229492188,
+      "learning_rate": 0.0005955867530887702,
+      "loss": 2.0175,
+      "step": 994
+    },
+    {
+      "epoch": 0.45621274644658416,
+      "grad_norm": 0.5683183670043945,
+      "learning_rate": 0.0005948576526505923,
+      "loss": 2.0991,
+      "step": 995
+    },
+    {
+      "epoch": 0.45667125171939477,
+      "grad_norm": 0.5283589363098145,
+      "learning_rate": 0.0005941283429210568,
+      "loss": 1.9816,
+      "step": 996
+    },
+    {
+      "epoch": 0.4571297569922054,
+      "grad_norm": 0.5492492914199829,
+      "learning_rate": 0.0005933988255092926,
+      "loss": 2.0796,
+      "step": 997
+    },
+    {
+      "epoch": 0.45758826226501603,
+      "grad_norm": 0.5823154449462891,
+      "learning_rate": 0.0005926691020248874,
+      "loss": 2.0628,
+      "step": 998
+    },
+    {
+      "epoch": 0.4580467675378267,
+      "grad_norm": 0.5661992430686951,
+      "learning_rate": 0.0005919391740778833,
+      "loss": 2.0318,
+      "step": 999
+    },
+    {
+      "epoch": 0.4585052728106373,
+      "grad_norm": 0.5616449117660522,
+      "learning_rate": 0.0005912090432787736,
+      "loss": 2.1314,
+      "step": 1000
+    },
+    {
+      "epoch": 0.45896377808344796,
+      "grad_norm": 0.5188776254653931,
+      "learning_rate": 0.000590478711238499,
+      "loss": 2.0272,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4594222833562586,
+      "grad_norm": 0.551785945892334,
+      "learning_rate": 0.0005897481795684446,
+      "loss": 2.0288,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4598807886290692,
+      "grad_norm": 0.5289655923843384,
+      "learning_rate": 0.0005890174498804355,
+      "loss": 2.0756,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4603392939018799,
+      "grad_norm": 0.5185794234275818,
+      "learning_rate": 0.0005882865237867339,
+      "loss": 2.026,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4607977991746905,
+      "grad_norm": 0.5245546698570251,
+      "learning_rate": 0.0005875554029000353,
+      "loss": 2.0549,
+      "step": 1005
+    },
+    {
+      "epoch": 0.46125630444750115,
+      "grad_norm": 0.5201156139373779,
+      "learning_rate": 0.0005868240888334653,
+      "loss": 2.0553,
+      "step": 1006
+    },
+    {
+      "epoch": 0.46171480972031176,
+      "grad_norm": 0.5486051440238953,
+      "learning_rate": 0.0005860925832005753,
+      "loss": 2.0901,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4621733149931224,
+      "grad_norm": 0.5827680826187134,
+      "learning_rate": 0.0005853608876153395,
+      "loss": 2.0793,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4626318202659331,
+      "grad_norm": 0.5337744951248169,
+      "learning_rate": 0.0005846290036921512,
+      "loss": 2.0167,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4630903255387437,
+      "grad_norm": 0.5507868528366089,
+      "learning_rate": 0.0005838969330458195,
+      "loss": 2.0383,
+      "step": 1010
+    },
+    {
+      "epoch": 0.46354883081155435,
+      "grad_norm": 0.5514399409294128,
+      "learning_rate": 0.0005831646772915651,
+      "loss": 2.0363,
+      "step": 1011
+    },
+    {
+      "epoch": 0.46400733608436495,
+      "grad_norm": 0.5788751244544983,
+      "learning_rate": 0.0005824322380450173,
+      "loss": 2.0838,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4644658413571756,
+      "grad_norm": 0.5801239013671875,
+      "learning_rate": 0.0005816996169222102,
+      "loss": 2.0702,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4649243466299862,
+      "grad_norm": 0.5164952278137207,
+      "learning_rate": 0.0005809668155395793,
+      "loss": 2.1026,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4653828519027969,
+      "grad_norm": 0.5412795543670654,
+      "learning_rate": 0.0005802338355139578,
+      "loss": 2.0171,
+      "step": 1015
+    },
+    {
+      "epoch": 0.46584135717560754,
+      "grad_norm": 0.5422810316085815,
+      "learning_rate": 0.0005795006784625728,
+      "loss": 2.1079,
+      "step": 1016
+    },
+    {
+      "epoch": 0.46629986244841815,
+      "grad_norm": 0.5823773741722107,
+      "learning_rate": 0.0005787673460030423,
+      "loss": 2.1081,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4667583677212288,
+      "grad_norm": 0.5180290937423706,
+      "learning_rate": 0.000578033839753371,
+      "loss": 2.1235,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4672168729940394,
+      "grad_norm": 0.5371957421302795,
+      "learning_rate": 0.0005773001613319476,
+      "loss": 1.9891,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4676753782668501,
+      "grad_norm": 0.5897166132926941,
+      "learning_rate": 0.00057656631235754,
+      "loss": 2.1223,
+      "step": 1020
+    },
+    {
+      "epoch": 0.46813388353966073,
+      "grad_norm": 0.5443727970123291,
+      "learning_rate": 0.0005758322944492929,
+      "loss": 2.0551,
+      "step": 1021
+    },
+    {
+      "epoch": 0.46859238881247134,
+      "grad_norm": 0.5617169141769409,
+      "learning_rate": 0.0005750981092267237,
+      "loss": 2.0706,
+      "step": 1022
+    },
+    {
+      "epoch": 0.469050894085282,
+      "grad_norm": 0.5548297762870789,
+      "learning_rate": 0.0005743637583097183,
+      "loss": 1.973,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4695093993580926,
+      "grad_norm": 0.5542665123939514,
+      "learning_rate": 0.0005736292433185291,
+      "loss": 1.9737,
+      "step": 1024
+    },
+    {
+      "epoch": 0.46996790463090327,
+      "grad_norm": 0.6357613205909729,
+      "learning_rate": 0.0005728945658737699,
+      "loss": 2.1335,
+      "step": 1025
+    },
+    {
+      "epoch": 0.47042640990371387,
+      "grad_norm": 0.5707666277885437,
+      "learning_rate": 0.0005721597275964133,
+      "loss": 2.0595,
+      "step": 1026
+    },
+    {
+      "epoch": 0.47088491517652453,
+      "grad_norm": 0.5593225955963135,
+      "learning_rate": 0.0005714247301077865,
+      "loss": 2.0641,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4713434204493352,
+      "grad_norm": 0.5901604294776917,
+      "learning_rate": 0.0005706895750295682,
+      "loss": 2.0899,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4718019257221458,
+      "grad_norm": 0.5472483038902283,
+      "learning_rate": 0.0005699542639837844,
+      "loss": 2.0637,
+      "step": 1029
+    },
+    {
+      "epoch": 0.47226043099495646,
+      "grad_norm": 0.5179505348205566,
+      "learning_rate": 0.0005692187985928055,
+      "loss": 2.1651,
+      "step": 1030
+    },
+    {
+      "epoch": 0.47271893626776706,
+      "grad_norm": 0.5251980423927307,
+      "learning_rate": 0.0005684831804793427,
+      "loss": 2.3032,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4731774415405777,
+      "grad_norm": 0.5827426314353943,
+      "learning_rate": 0.0005677474112664438,
+      "loss": 2.032,
+      "step": 1032
+    },
+    {
+      "epoch": 0.47363594681338833,
+      "grad_norm": 0.6011113524436951,
+      "learning_rate": 0.0005670114925774899,
+      "loss": 2.1042,
+      "step": 1033
+    },
+    {
+      "epoch": 0.474094452086199,
+      "grad_norm": 0.5491259694099426,
+      "learning_rate": 0.0005662754260361924,
+      "loss": 2.0919,
+      "step": 1034
+    },
+    {
+      "epoch": 0.47455295735900965,
+      "grad_norm": 0.5617481470108032,
+      "learning_rate": 0.0005655392132665884,
+      "loss": 2.0261,
+      "step": 1035
+    },
+    {
+      "epoch": 0.47501146263182026,
+      "grad_norm": 0.6067972183227539,
+      "learning_rate": 0.000564802855893038,
+      "loss": 2.0586,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4754699679046309,
+      "grad_norm": 0.5716037154197693,
+      "learning_rate": 0.0005640663555402198,
+      "loss": 2.1535,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4759284731774415,
+      "grad_norm": 0.5653804540634155,
+      "learning_rate": 0.0005633297138331285,
+      "loss": 2.1112,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4763869784502522,
+      "grad_norm": 0.5448452234268188,
+      "learning_rate": 0.0005625929323970705,
+      "loss": 2.214,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4768454837230628,
+      "grad_norm": 0.5342665910720825,
+      "learning_rate": 0.0005618560128576603,
+      "loss": 2.0379,
+      "step": 1040
+    },
+    {
+      "epoch": 0.47730398899587345,
+      "grad_norm": 0.5197364687919617,
+      "learning_rate": 0.0005611189568408173,
+      "loss": 2.0699,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4777624942686841,
+      "grad_norm": 0.5099651217460632,
+      "learning_rate": 0.0005603817659727619,
+      "loss": 2.0109,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4782209995414947,
+      "grad_norm": 0.5607723593711853,
+      "learning_rate": 0.0005596444418800121,
+      "loss": 2.0286,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4786795048143054,
+      "grad_norm": 0.5326158404350281,
+      "learning_rate": 0.0005589069861893798,
+      "loss": 2.0525,
+      "step": 1044
+    },
+    {
+      "epoch": 0.479138010087116,
+      "grad_norm": 0.5216630101203918,
+      "learning_rate": 0.0005581694005279673,
+      "loss": 2.0489,
+      "step": 1045
+    },
+    {
+      "epoch": 0.47959651535992665,
+      "grad_norm": 0.5481088161468506,
+      "learning_rate": 0.0005574316865231637,
+      "loss": 1.9948,
+      "step": 1046
+    },
+    {
+      "epoch": 0.48005502063273725,
+      "grad_norm": 0.5221377611160278,
+      "learning_rate": 0.0005566938458026411,
+      "loss": 2.0385,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4805135259055479,
+      "grad_norm": 0.5405771732330322,
+      "learning_rate": 0.0005559558799943514,
+      "loss": 1.9893,
+      "step": 1048
+    },
+    {
+      "epoch": 0.48097203117835857,
+      "grad_norm": 0.5451287627220154,
+      "learning_rate": 0.0005552177907265223,
+      "loss": 1.8918,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4814305364511692,
+      "grad_norm": 0.5571972131729126,
+      "learning_rate": 0.000554479579627654,
+      "loss": 2.0191,
+      "step": 1050
+    },
+    {
+      "epoch": 0.48188904172397984,
+      "grad_norm": 0.5704662799835205,
+      "learning_rate": 0.0005537412483265157,
+      "loss": 2.0803,
+      "step": 1051
+    },
+    {
+      "epoch": 0.48234754699679044,
+      "grad_norm": 0.56961989402771,
+      "learning_rate": 0.0005530027984521413,
+      "loss": 2.0787,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4828060522696011,
+      "grad_norm": 0.5583857893943787,
+      "learning_rate": 0.0005522642316338268,
+      "loss": 1.9951,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4832645575424117,
+      "grad_norm": 0.5495345592498779,
+      "learning_rate": 0.0005515255495011259,
+      "loss": 2.02,
+      "step": 1054
+    },
+    {
+      "epoch": 0.48372306281522237,
+      "grad_norm": 0.5690191388130188,
+      "learning_rate": 0.0005507867536838472,
+      "loss": 2.0444,
+      "step": 1055
+    },
+    {
+      "epoch": 0.48418156808803303,
+      "grad_norm": 0.5902395844459534,
+      "learning_rate": 0.0005500478458120492,
+      "loss": 1.9998,
+      "step": 1056
+    },
+    {
+      "epoch": 0.48464007336084364,
+      "grad_norm": 0.5545502305030823,
+      "learning_rate": 0.0005493088275160387,
+      "loss": 2.0968,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4850985786336543,
+      "grad_norm": 0.5170363187789917,
+      "learning_rate": 0.0005485697004263657,
+      "loss": 2.0557,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4855570839064649,
+      "grad_norm": 0.5218999981880188,
+      "learning_rate": 0.0005478304661738199,
+      "loss": 2.0938,
+      "step": 1059
+    },
+    {
+      "epoch": 0.48601558917927556,
+      "grad_norm": 0.6168057918548584,
+      "learning_rate": 0.0005470911263894279,
+      "loss": 2.0053,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4864740944520862,
+      "grad_norm": 0.5872636437416077,
+      "learning_rate": 0.0005463516827044491,
+      "loss": 2.0836,
+      "step": 1061
+    },
+    {
+      "epoch": 0.48693259972489683,
+      "grad_norm": 0.5243323445320129,
+      "learning_rate": 0.000545612136750372,
+      "loss": 2.0047,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4873911049977075,
+      "grad_norm": 0.5484859347343445,
+      "learning_rate": 0.0005448724901589107,
+      "loss": 2.1724,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4878496102705181,
+      "grad_norm": 0.5737951397895813,
+      "learning_rate": 0.0005441327445620014,
+      "loss": 2.0524,
+      "step": 1064
+    },
+    {
+      "epoch": 0.48830811554332876,
+      "grad_norm": 0.5733909010887146,
+      "learning_rate": 0.0005433929015917988,
+      "loss": 2.016,
+      "step": 1065
+    },
+    {
+      "epoch": 0.48876662081613936,
+      "grad_norm": 0.5204495787620544,
+      "learning_rate": 0.0005426529628806724,
+      "loss": 2.0078,
+      "step": 1066
+    },
+    {
+      "epoch": 0.48922512608895,
+      "grad_norm": 0.5341218709945679,
+      "learning_rate": 0.0005419129300612029,
+      "loss": 2.0631,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4896836313617607,
+      "grad_norm": 0.5687958002090454,
+      "learning_rate": 0.000541172804766179,
+      "loss": 2.1547,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4901421366345713,
+      "grad_norm": 0.5223873853683472,
+      "learning_rate": 0.0005404325886285927,
+      "loss": 1.975,
+      "step": 1069
+    },
+    {
+      "epoch": 0.49060064190738195,
+      "grad_norm": 0.5557146668434143,
+      "learning_rate": 0.000539692283281637,
+      "loss": 2.1069,
+      "step": 1070
+    },
+    {
+      "epoch": 0.49105914718019256,
+      "grad_norm": 0.5473342537879944,
+      "learning_rate": 0.0005389518903587017,
+      "loss": 2.0686,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4915176524530032,
+      "grad_norm": 0.5465348958969116,
+      "learning_rate": 0.0005382114114933695,
+      "loss": 2.0073,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4919761577258138,
+      "grad_norm": 0.5557299852371216,
+      "learning_rate": 0.0005374708483194132,
+      "loss": 2.0225,
+      "step": 1073
+    },
+    {
+      "epoch": 0.4924346629986245,
+      "grad_norm": 0.5395819544792175,
+      "learning_rate": 0.000536730202470791,
+      "loss": 2.0192,
+      "step": 1074
+    },
+    {
+      "epoch": 0.49289316827143514,
+      "grad_norm": 0.6064382791519165,
+      "learning_rate": 0.0005359894755816443,
+      "loss": 2.0105,
+      "step": 1075
+    },
+    {
+      "epoch": 0.49335167354424575,
+      "grad_norm": 0.5188748240470886,
+      "learning_rate": 0.0005352486692862926,
+      "loss": 2.0584,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4938101788170564,
+      "grad_norm": 0.5620332956314087,
+      "learning_rate": 0.0005345077852192307,
+      "loss": 2.0943,
+      "step": 1077
+    },
+    {
+      "epoch": 0.494268684089867,
+      "grad_norm": 0.558038592338562,
+      "learning_rate": 0.0005337668250151254,
+      "loss": 2.0331,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4947271893626777,
+      "grad_norm": 0.5156983733177185,
+      "learning_rate": 0.0005330257903088111,
+      "loss": 2.0629,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4951856946354883,
+      "grad_norm": 0.5393120050430298,
+      "learning_rate": 0.000532284682735287,
+      "loss": 1.9769,
+      "step": 1080
+    },
+    {
+      "epoch": 0.49564419990829894,
+      "grad_norm": 0.5530503392219543,
+      "learning_rate": 0.0005315435039297124,
+      "loss": 2.0694,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4961027051811096,
+      "grad_norm": 0.5521067380905151,
+      "learning_rate": 0.0005308022555274046,
+      "loss": 2.0111,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4965612104539202,
+      "grad_norm": 0.5657058358192444,
+      "learning_rate": 0.0005300609391638336,
+      "loss": 2.0232,
+      "step": 1083
+    },
+    {
+      "epoch": 0.49701971572673087,
+      "grad_norm": 0.5275099873542786,
+      "learning_rate": 0.0005293195564746201,
+      "loss": 2.0189,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4974782209995415,
+      "grad_norm": 0.5172412991523743,
+      "learning_rate": 0.0005285781090955304,
+      "loss": 1.9818,
+      "step": 1085
+    },
+    {
+      "epoch": 0.49793672627235214,
+      "grad_norm": 0.5378063321113586,
+      "learning_rate": 0.0005278365986624743,
+      "loss": 2.0063,
+      "step": 1086
+    },
+    {
+      "epoch": 0.49839523154516274,
+      "grad_norm": 0.5514046549797058,
+      "learning_rate": 0.0005270950268115001,
+      "loss": 1.9581,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4988537368179734,
+      "grad_norm": 0.5331932902336121,
+      "learning_rate": 0.0005263533951787919,
+      "loss": 2.1113,
+      "step": 1088
+    },
+    {
+      "epoch": 0.49931224209078406,
+      "grad_norm": 0.5486401319503784,
+      "learning_rate": 0.000525611705400666,
+      "loss": 2.0704,
+      "step": 1089
+    },
+    {
+      "epoch": 0.49977074736359467,
+      "grad_norm": 0.5018776059150696,
+      "learning_rate": 0.0005248699591135664,
+      "loss": 2.0635,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5002292526364053,
+      "grad_norm": 0.587924599647522,
+      "learning_rate": 0.0005241281579540618,
+      "loss": 1.9991,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5006877579092159,
+      "grad_norm": 0.5879606604576111,
+      "learning_rate": 0.0005233863035588427,
+      "loss": 2.1343,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5011462631820266,
+      "grad_norm": 0.5524023175239563,
+      "learning_rate": 0.0005226443975647161,
+      "loss": 2.1166,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5016047684548373,
+      "grad_norm": 0.537943422794342,
+      "learning_rate": 0.0005219024416086036,
+      "loss": 2.034,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5020632737276479,
+      "grad_norm": 0.51746666431427,
+      "learning_rate": 0.0005211604373275366,
+      "loss": 2.084,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5025217790004585,
+      "grad_norm": 0.5621870756149292,
+      "learning_rate": 0.0005204183863586533,
+      "loss": 1.9873,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5029802842732691,
+      "grad_norm": 0.5723305344581604,
+      "learning_rate": 0.0005196762903391951,
+      "loss": 2.0642,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5034387895460798,
+      "grad_norm": 0.5384237170219421,
+      "learning_rate": 0.0005189341509065023,
+      "loss": 2.0342,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5038972948188905,
+      "grad_norm": 0.515001118183136,
+      "learning_rate": 0.0005181919696980112,
+      "loss": 2.0944,
+      "step": 1099
+    },
+    {
+      "epoch": 0.504355800091701,
+      "grad_norm": 0.5616503953933716,
+      "learning_rate": 0.0005174497483512506,
+      "loss": 2.0628,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5048143053645117,
+      "grad_norm": 0.5357555747032166,
+      "learning_rate": 0.0005167074885038374,
+      "loss": 1.9591,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5052728106373223,
+      "grad_norm": 0.526648223400116,
+      "learning_rate": 0.0005159651917934735,
+      "loss": 2.0143,
+      "step": 1102
+    },
+    {
+      "epoch": 0.505731315910133,
+      "grad_norm": 0.5753529071807861,
+      "learning_rate": 0.0005152228598579428,
+      "loss": 2.1238,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5061898211829436,
+      "grad_norm": 0.5752229690551758,
+      "learning_rate": 0.000514480494335106,
+      "loss": 2.0204,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5066483264557542,
+      "grad_norm": 0.5407384037971497,
+      "learning_rate": 0.0005137380968628983,
+      "loss": 2.0274,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5071068317285649,
+      "grad_norm": 0.5850804448127747,
+      "learning_rate": 0.0005129956690793255,
+      "loss": 2.0305,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5075653370013755,
+      "grad_norm": 0.5756503343582153,
+      "learning_rate": 0.0005122532126224601,
+      "loss": 2.1214,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5080238422741862,
+      "grad_norm": 0.5153061151504517,
+      "learning_rate": 0.0005115107291304378,
+      "loss": 2.0452,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5084823475469968,
+      "grad_norm": 0.5745264291763306,
+      "learning_rate": 0.0005107682202414544,
+      "loss": 2.0417,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5089408528198074,
+      "grad_norm": 0.5371508002281189,
+      "learning_rate": 0.0005100256875937613,
+      "loss": 1.9995,
+      "step": 1110
+    },
+    {
+      "epoch": 0.509399358092618,
+      "grad_norm": 0.5526379346847534,
+      "learning_rate": 0.0005092831328256625,
+      "loss": 2.1066,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5098578633654287,
+      "grad_norm": 0.5020286440849304,
+      "learning_rate": 0.0005085405575755105,
+      "loss": 2.0549,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5103163686382394,
+      "grad_norm": 0.5453764200210571,
+      "learning_rate": 0.0005077979634817034,
+      "loss": 1.9608,
+      "step": 1113
+    },
+    {
+      "epoch": 0.51077487391105,
+      "grad_norm": 0.5326483845710754,
+      "learning_rate": 0.0005070553521826808,
+      "loss": 1.9919,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5112333791838606,
+      "grad_norm": 0.535541832447052,
+      "learning_rate": 0.00050631272531692,
+      "loss": 2.0946,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5116918844566712,
+      "grad_norm": 0.5388101935386658,
+      "learning_rate": 0.0005055700845229327,
+      "loss": 2.0818,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5121503897294819,
+      "grad_norm": 0.520546555519104,
+      "learning_rate": 0.000504827431439262,
+      "loss": 2.0506,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5126088950022926,
+      "grad_norm": 0.5576258897781372,
+      "learning_rate": 0.000504084767704477,
+      "loss": 2.067,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5130674002751031,
+      "grad_norm": 0.5915591716766357,
+      "learning_rate": 0.0005033420949571712,
+      "loss": 1.9967,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5135259055479138,
+      "grad_norm": 0.5337035655975342,
+      "learning_rate": 0.0005025994148359574,
+      "loss": 2.0609,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5139844108207244,
+      "grad_norm": 0.5410395264625549,
+      "learning_rate": 0.0005018567289794651,
+      "loss": 1.9331,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5144429160935351,
+      "grad_norm": 0.6455917954444885,
+      "learning_rate": 0.0005011140390263362,
+      "loss": 2.1111,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5149014213663458,
+      "grad_norm": 0.5802803039550781,
+      "learning_rate": 0.0005003713466152218,
+      "loss": 2.0191,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5153599266391563,
+      "grad_norm": 0.5285162925720215,
+      "learning_rate": 0.0004996286533847783,
+      "loss": 2.0619,
+      "step": 1124
+    },
+    {
+      "epoch": 0.515818431911967,
+      "grad_norm": 0.5374143123626709,
+      "learning_rate": 0.000498885960973664,
+      "loss": 2.035,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5162769371847776,
+      "grad_norm": 0.5920125246047974,
+      "learning_rate": 0.000498143271020535,
+      "loss": 2.0121,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5167354424575883,
+      "grad_norm": 0.5320558547973633,
+      "learning_rate": 0.0004974005851640428,
+      "loss": 2.0602,
+      "step": 1127
+    },
+    {
+      "epoch": 0.517193947730399,
+      "grad_norm": 0.5635319948196411,
+      "learning_rate": 0.000496657905042829,
+      "loss": 2.0464,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5176524530032095,
+      "grad_norm": 0.5621376037597656,
+      "learning_rate": 0.0004959152322955232,
+      "loss": 2.0218,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5181109582760202,
+      "grad_norm": 0.5368046164512634,
+      "learning_rate": 0.0004951725685607382,
+      "loss": 2.1621,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5185694635488308,
+      "grad_norm": 0.5438060760498047,
+      "learning_rate": 0.0004944299154770673,
+      "loss": 2.0,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5190279688216415,
+      "grad_norm": 0.549072265625,
+      "learning_rate": 0.0004936872746830802,
+      "loss": 2.0359,
+      "step": 1132
+    },
+    {
+      "epoch": 0.519486474094452,
+      "grad_norm": 0.5452683568000793,
+      "learning_rate": 0.0004929446478173195,
+      "loss": 2.0599,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5199449793672627,
+      "grad_norm": 0.5487686395645142,
+      "learning_rate": 0.0004922020365182968,
+      "loss": 2.0053,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5204034846400734,
+      "grad_norm": 0.5297566056251526,
+      "learning_rate": 0.0004914594424244897,
+      "loss": 2.0305,
+      "step": 1135
+    },
+    {
+      "epoch": 0.520861989912884,
+      "grad_norm": 0.5251104235649109,
+      "learning_rate": 0.0004907168671743376,
+      "loss": 2.0217,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5213204951856947,
+      "grad_norm": 0.5246418118476868,
+      "learning_rate": 0.0004899743124062387,
+      "loss": 1.9859,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5217790004585052,
+      "grad_norm": 0.5447359681129456,
+      "learning_rate": 0.0004892317797585456,
+      "loss": 1.9776,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5222375057313159,
+      "grad_norm": 0.569364070892334,
+      "learning_rate": 0.0004884892708695623,
+      "loss": 1.9535,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5226960110041265,
+      "grad_norm": 0.5333812832832336,
+      "learning_rate": 0.0004877467873775402,
+      "loss": 2.0293,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5231545162769372,
+      "grad_norm": 0.5415140390396118,
+      "learning_rate": 0.00048700433092067473,
+      "loss": 2.0712,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5236130215497479,
+      "grad_norm": 0.5224328637123108,
+      "learning_rate": 0.0004862619031371019,
+      "loss": 2.0296,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5240715268225584,
+      "grad_norm": 0.533515989780426,
+      "learning_rate": 0.0004855195056648942,
+      "loss": 2.0055,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5245300320953691,
+      "grad_norm": 0.5211889743804932,
+      "learning_rate": 0.00048477714014205734,
+      "loss": 2.0241,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5249885373681797,
+      "grad_norm": 0.5496058464050293,
+      "learning_rate": 0.00048403480820652644,
+      "loss": 2.1108,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5254470426409904,
+      "grad_norm": 0.5099970698356628,
+      "learning_rate": 0.0004832925114961629,
+      "loss": 2.0433,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5259055479138011,
+      "grad_norm": 0.5410376787185669,
+      "learning_rate": 0.0004825502516487497,
+      "loss": 2.048,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5263640531866116,
+      "grad_norm": 0.546927273273468,
+      "learning_rate": 0.00048180803030198896,
+      "loss": 2.1379,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5268225584594223,
+      "grad_norm": 0.5357005000114441,
+      "learning_rate": 0.0004810658490934979,
+      "loss": 2.0705,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5272810637322329,
+      "grad_norm": 0.5324466228485107,
+      "learning_rate": 0.000480323709660805,
+      "loss": 2.1019,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5277395690050436,
+      "grad_norm": 0.529270350933075,
+      "learning_rate": 0.0004795816136413467,
+      "loss": 1.9589,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5281980742778541,
+      "grad_norm": 0.543827474117279,
+      "learning_rate": 0.00047883956267246353,
+      "loss": 2.0695,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5286565795506648,
+      "grad_norm": 0.5250212550163269,
+      "learning_rate": 0.00047809755839139657,
+      "loss": 2.0611,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5291150848234755,
+      "grad_norm": 0.5148019790649414,
+      "learning_rate": 0.0004773556024352841,
+      "loss": 2.0573,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5295735900962861,
+      "grad_norm": 0.5067748427391052,
+      "learning_rate": 0.00047661369644115754,
+      "loss": 1.9179,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5300320953690968,
+      "grad_norm": 0.5363000631332397,
+      "learning_rate": 0.0004758718420459383,
+      "loss": 1.9821,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5304906006419073,
+      "grad_norm": 0.5441141128540039,
+      "learning_rate": 0.0004751300408864339,
+      "loss": 1.961,
+      "step": 1157
+    },
+    {
+      "epoch": 0.530949105914718,
+      "grad_norm": 0.522655725479126,
+      "learning_rate": 0.00047438829459933414,
+      "loss": 2.0333,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5314076111875287,
+      "grad_norm": 0.5663999319076538,
+      "learning_rate": 0.0004736466048212082,
+      "loss": 2.0938,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5318661164603393,
+      "grad_norm": 0.5321094393730164,
+      "learning_rate": 0.0004729049731885002,
+      "loss": 2.017,
+      "step": 1160
+    },
+    {
+      "epoch": 0.53232462173315,
+      "grad_norm": 0.5095037221908569,
+      "learning_rate": 0.000472163401337526,
+      "loss": 2.0133,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5327831270059605,
+      "grad_norm": 0.5180181264877319,
+      "learning_rate": 0.00047142189090446985,
+      "loss": 1.9738,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5332416322787712,
+      "grad_norm": 0.5090748071670532,
+      "learning_rate": 0.0004706804435253802,
+      "loss": 2.1304,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5337001375515819,
+      "grad_norm": 0.5436115264892578,
+      "learning_rate": 0.0004699390608361665,
+      "loss": 2.1919,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5341586428243925,
+      "grad_norm": 0.5561127066612244,
+      "learning_rate": 0.0004691977444725955,
+      "loss": 2.0316,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5346171480972031,
+      "grad_norm": 0.5897061228752136,
+      "learning_rate": 0.0004684564960702877,
+      "loss": 2.0529,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5350756533700137,
+      "grad_norm": 0.5204058885574341,
+      "learning_rate": 0.0004677153172647131,
+      "loss": 2.0377,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5355341586428244,
+      "grad_norm": 0.5606356263160706,
+      "learning_rate": 0.00046697420969118894,
+      "loss": 2.1047,
+      "step": 1168
+    },
+    {
+      "epoch": 0.535992663915635,
+      "grad_norm": 0.5508114099502563,
+      "learning_rate": 0.00046623317498487466,
+      "loss": 2.0039,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5364511691884457,
+      "grad_norm": 0.5592308640480042,
+      "learning_rate": 0.0004654922147807694,
+      "loss": 2.1272,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5369096744612563,
+      "grad_norm": 0.5230913758277893,
+      "learning_rate": 0.00046475133071370757,
+      "loss": 1.9847,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5373681797340669,
+      "grad_norm": 0.5270746946334839,
+      "learning_rate": 0.00046401052441835574,
+      "loss": 2.0088,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5378266850068776,
+      "grad_norm": 0.5212060213088989,
+      "learning_rate": 0.000463269797529209,
+      "loss": 1.9261,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5382851902796882,
+      "grad_norm": 0.5368192791938782,
+      "learning_rate": 0.00046252915168058697,
+      "loss": 2.0636,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5387436955524989,
+      "grad_norm": 0.529940664768219,
+      "learning_rate": 0.0004617885885066305,
+      "loss": 1.9799,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5392022008253095,
+      "grad_norm": 0.5496656894683838,
+      "learning_rate": 0.0004610481096412984,
+      "loss": 2.034,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5396607060981201,
+      "grad_norm": 0.5195562243461609,
+      "learning_rate": 0.000460307716718363,
+      "loss": 2.0038,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5401192113709308,
+      "grad_norm": 0.5596598386764526,
+      "learning_rate": 0.0004595674113714074,
+      "loss": 2.0354,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5405777166437414,
+      "grad_norm": 0.5272588133811951,
+      "learning_rate": 0.0004588271952338212,
+      "loss": 2.0766,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5410362219165521,
+      "grad_norm": 0.5344094038009644,
+      "learning_rate": 0.00045808706993879714,
+      "loss": 2.1139,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5414947271893626,
+      "grad_norm": 0.5276915431022644,
+      "learning_rate": 0.00045734703711932767,
+      "loss": 2.0291,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5419532324621733,
+      "grad_norm": 0.5204179286956787,
+      "learning_rate": 0.0004566070984082013,
+      "loss": 1.9406,
+      "step": 1182
+    },
+    {
+      "epoch": 0.542411737734984,
+      "grad_norm": 0.518652081489563,
+      "learning_rate": 0.00045586725543799865,
+      "loss": 1.9673,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5428702430077946,
+      "grad_norm": 0.5278819799423218,
+      "learning_rate": 0.00045512750984108937,
+      "loss": 2.0743,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5433287482806052,
+      "grad_norm": 0.49915292859077454,
+      "learning_rate": 0.000454387863249628,
+      "loss": 2.0902,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5437872535534158,
+      "grad_norm": 0.5552077889442444,
+      "learning_rate": 0.00045364831729555096,
+      "loss": 2.0218,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5442457588262265,
+      "grad_norm": 0.5096197128295898,
+      "learning_rate": 0.0004529088736105721,
+      "loss": 2.0536,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5447042640990372,
+      "grad_norm": 0.5648066401481628,
+      "learning_rate": 0.0004521695338261802,
+      "loss": 2.0765,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5451627693718478,
+      "grad_norm": 0.5497797727584839,
+      "learning_rate": 0.0004514302995736344,
+      "loss": 2.0581,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5456212746446584,
+      "grad_norm": 0.50295490026474,
+      "learning_rate": 0.0004506911724839613,
+      "loss": 2.0013,
+      "step": 1190
+    },
+    {
+      "epoch": 0.546079779917469,
+      "grad_norm": 0.529457151889801,
+      "learning_rate": 0.0004499521541879508,
+      "loss": 1.9561,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5465382851902797,
+      "grad_norm": 0.5721673369407654,
+      "learning_rate": 0.00044921324631615303,
+      "loss": 2.0193,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5469967904630904,
+      "grad_norm": 0.5277779698371887,
+      "learning_rate": 0.0004484744504988742,
+      "loss": 1.9882,
+      "step": 1193
+    },
+    {
+      "epoch": 0.547455295735901,
+      "grad_norm": 0.5685149431228638,
+      "learning_rate": 0.00044773576836617336,
+      "loss": 1.99,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5479138010087116,
+      "grad_norm": 0.6087177395820618,
+      "learning_rate": 0.0004469972015478588,
+      "loss": 2.0474,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5483723062815222,
+      "grad_norm": 0.5323075652122498,
+      "learning_rate": 0.0004462587516734844,
+      "loss": 1.9989,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5488308115543329,
+      "grad_norm": 0.5347911715507507,
+      "learning_rate": 0.00044552042037234596,
+      "loss": 2.0639,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5492893168271435,
+      "grad_norm": 0.5470256805419922,
+      "learning_rate": 0.00044478220927347774,
+      "loss": 2.0469,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5497478220999541,
+      "grad_norm": 0.5137409567832947,
+      "learning_rate": 0.00044404412000564875,
+      "loss": 1.9941,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5502063273727648,
+      "grad_norm": 0.5459279417991638,
+      "learning_rate": 0.000443306154197359,
+      "loss": 2.0278,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5506648326455754,
+      "grad_norm": 0.5826395750045776,
+      "learning_rate": 0.00044256831347683646,
+      "loss": 2.0005,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5511233379183861,
+      "grad_norm": 0.5612067580223083,
+      "learning_rate": 0.0004418305994720328,
+      "loss": 2.005,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5515818431911967,
+      "grad_norm": 0.5044583082199097,
+      "learning_rate": 0.0004410930138106203,
+      "loss": 2.0082,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5520403484640073,
+      "grad_norm": 0.5432026982307434,
+      "learning_rate": 0.000440355558119988,
+      "loss": 2.1139,
+      "step": 1204
+    },
+    {
+      "epoch": 0.552498853736818,
+      "grad_norm": 0.5684635043144226,
+      "learning_rate": 0.00043961823402723814,
+      "loss": 2.0273,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5529573590096286,
+      "grad_norm": 0.575829267501831,
+      "learning_rate": 0.0004388810431591829,
+      "loss": 2.1018,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5534158642824393,
+      "grad_norm": 0.590873658657074,
+      "learning_rate": 0.0004381439871423398,
+      "loss": 1.9704,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5538743695552499,
+      "grad_norm": 0.5433682799339294,
+      "learning_rate": 0.00043740706760292966,
+      "loss": 2.0284,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5543328748280605,
+      "grad_norm": 0.5501098036766052,
+      "learning_rate": 0.0004366702861668716,
+      "loss": 2.0562,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5547913801008711,
+      "grad_norm": 0.5647679567337036,
+      "learning_rate": 0.00043593364445978036,
+      "loss": 2.0693,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5552498853736818,
+      "grad_norm": 0.5750144124031067,
+      "learning_rate": 0.0004351971441069622,
+      "loss": 1.9932,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5557083906464925,
+      "grad_norm": 0.5219569206237793,
+      "learning_rate": 0.0004344607867334116,
+      "loss": 1.9813,
+      "step": 1212
+    },
+    {
+      "epoch": 0.556166895919303,
+      "grad_norm": 0.5695562958717346,
+      "learning_rate": 0.00043372457396380766,
+      "loss": 1.9594,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5566254011921137,
+      "grad_norm": 0.6164484620094299,
+      "learning_rate": 0.00043298850742251013,
+      "loss": 2.1527,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5570839064649243,
+      "grad_norm": 0.5321323275566101,
+      "learning_rate": 0.0004322525887335563,
+      "loss": 2.0055,
+      "step": 1215
+    },
+    {
+      "epoch": 0.557542411737735,
+      "grad_norm": 0.5484564900398254,
+      "learning_rate": 0.00043151681952065734,
+      "loss": 2.055,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5580009170105457,
+      "grad_norm": 0.5348613262176514,
+      "learning_rate": 0.00043078120140719456,
+      "loss": 2.0968,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5584594222833562,
+      "grad_norm": 0.5568328499794006,
+      "learning_rate": 0.0004300457360162158,
+      "loss": 2.0548,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5589179275561669,
+      "grad_norm": 0.5506913661956787,
+      "learning_rate": 0.0004293104249704319,
+      "loss": 2.0032,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5593764328289775,
+      "grad_norm": 0.5490648746490479,
+      "learning_rate": 0.00042857526989221355,
+      "loss": 2.0745,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5598349381017882,
+      "grad_norm": 0.5258785486221313,
+      "learning_rate": 0.00042784027240358674,
+      "loss": 2.0471,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5602934433745989,
+      "grad_norm": 0.5345601439476013,
+      "learning_rate": 0.0004271054341262301,
+      "loss": 2.0148,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5607519486474094,
+      "grad_norm": 0.517360508441925,
+      "learning_rate": 0.000426370756681471,
+      "loss": 2.0122,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5612104539202201,
+      "grad_norm": 0.5245397090911865,
+      "learning_rate": 0.0004256362416902817,
+      "loss": 2.0249,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5616689591930307,
+      "grad_norm": 0.5636434555053711,
+      "learning_rate": 0.00042490189077327637,
+      "loss": 2.0428,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5621274644658414,
+      "grad_norm": 0.5474424958229065,
+      "learning_rate": 0.00042416770555070703,
+      "loss": 2.0179,
+      "step": 1226
+    },
+    {
+      "epoch": 0.562585969738652,
+      "grad_norm": 0.5320186614990234,
+      "learning_rate": 0.00042343368764246,
+      "loss": 1.9481,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5630444750114626,
+      "grad_norm": 0.5331994891166687,
+      "learning_rate": 0.0004226998386680524,
+      "loss": 2.0862,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5635029802842733,
+      "grad_norm": 0.5148548483848572,
+      "learning_rate": 0.000421966160246629,
+      "loss": 2.1413,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5639614855570839,
+      "grad_norm": 0.564822256565094,
+      "learning_rate": 0.00042123265399695783,
+      "loss": 2.0379,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5644199908298946,
+      "grad_norm": 0.5515744090080261,
+      "learning_rate": 0.0004204993215374273,
+      "loss": 2.1457,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5648784961027051,
+      "grad_norm": 0.5053184628486633,
+      "learning_rate": 0.00041976616448604226,
+      "loss": 2.049,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5653370013755158,
+      "grad_norm": 0.516232430934906,
+      "learning_rate": 0.00041903318446042076,
+      "loss": 2.0228,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5657955066483265,
+      "grad_norm": 0.5145546793937683,
+      "learning_rate": 0.00041830038307778984,
+      "loss": 2.0953,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5662540119211371,
+      "grad_norm": 0.5331826210021973,
+      "learning_rate": 0.0004175677619549828,
+      "loss": 2.1206,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5667125171939478,
+      "grad_norm": 0.5112865567207336,
+      "learning_rate": 0.000416835322708435,
+      "loss": 1.9387,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5671710224667583,
+      "grad_norm": 0.5202925801277161,
+      "learning_rate": 0.00041610306695418056,
+      "loss": 1.9846,
+      "step": 1237
+    },
+    {
+      "epoch": 0.567629527739569,
+      "grad_norm": 0.5414956212043762,
+      "learning_rate": 0.0004153709963078488,
+      "loss": 2.0719,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5680880330123796,
+      "grad_norm": 0.528677225112915,
+      "learning_rate": 0.0004146391123846606,
+      "loss": 2.0362,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5685465382851903,
+      "grad_norm": 0.5216994881629944,
+      "learning_rate": 0.0004139074167994249,
+      "loss": 2.0044,
+      "step": 1240
+    },
+    {
+      "epoch": 0.569005043558001,
+      "grad_norm": 0.53104567527771,
+      "learning_rate": 0.00041317591116653486,
+      "loss": 2.0859,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5694635488308115,
+      "grad_norm": 0.5055778622627258,
+      "learning_rate": 0.0004124445970999648,
+      "loss": 2.0374,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5699220541036222,
+      "grad_norm": 0.546288013458252,
+      "learning_rate": 0.00041171347621326627,
+      "loss": 2.057,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5703805593764328,
+      "grad_norm": 0.5225109457969666,
+      "learning_rate": 0.00041098255011956465,
+      "loss": 2.0499,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5708390646492435,
+      "grad_norm": 0.538680911064148,
+      "learning_rate": 0.00041025182043155547,
+      "loss": 2.0718,
+      "step": 1245
+    },
+    {
+      "epoch": 0.571297569922054,
+      "grad_norm": 0.5589045882225037,
+      "learning_rate": 0.000409521288761501,
+      "loss": 2.0904,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5717560751948647,
+      "grad_norm": 0.5503789782524109,
+      "learning_rate": 0.00040879095672122646,
+      "loss": 2.0326,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5722145804676754,
+      "grad_norm": 0.5759222507476807,
+      "learning_rate": 0.0004080608259221167,
+      "loss": 2.0259,
+      "step": 1248
+    },
+    {
+      "epoch": 0.572673085740486,
+      "grad_norm": 0.5171306729316711,
+      "learning_rate": 0.0004073308979751126,
+      "loss": 1.9712,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5731315910132967,
+      "grad_norm": 0.5442126989364624,
+      "learning_rate": 0.0004066011744907074,
+      "loss": 1.9639,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5735900962861072,
+      "grad_norm": 0.5553286671638489,
+      "learning_rate": 0.00040587165707894326,
+      "loss": 2.0544,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5740486015589179,
+      "grad_norm": 0.5304802656173706,
+      "learning_rate": 0.0004051423473494076,
+      "loss": 1.9764,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5745071068317286,
+      "grad_norm": 0.5310506224632263,
+      "learning_rate": 0.0004044132469112299,
+      "loss": 2.0093,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5749656121045392,
+      "grad_norm": 0.5661071538925171,
+      "learning_rate": 0.00040368435737307733,
+      "loss": 2.0926,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5754241173773499,
+      "grad_norm": 0.5356221199035645,
+      "learning_rate": 0.00040295568034315224,
+      "loss": 1.9488,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5758826226501604,
+      "grad_norm": 0.5154054760932922,
+      "learning_rate": 0.0004022272174291878,
+      "loss": 2.1181,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5763411279229711,
+      "grad_norm": 0.5422287583351135,
+      "learning_rate": 0.0004014989702384449,
+      "loss": 1.9568,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5767996331957818,
+      "grad_norm": 0.5153090953826904,
+      "learning_rate": 0.00040077094037770843,
+      "loss": 2.0322,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5772581384685924,
+      "grad_norm": 0.5375383496284485,
+      "learning_rate": 0.0004000431294532838,
+      "loss": 2.0133,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5777166437414031,
+      "grad_norm": 0.5344530344009399,
+      "learning_rate": 0.0003993155390709935,
+      "loss": 2.0474,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5781751490142136,
+      "grad_norm": 0.5306925773620605,
+      "learning_rate": 0.0003985881708361729,
+      "loss": 2.0742,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5786336542870243,
+      "grad_norm": 0.516596257686615,
+      "learning_rate": 0.00039786102635366784,
+      "loss": 1.9892,
+      "step": 1262
+    },
+    {
+      "epoch": 0.579092159559835,
+      "grad_norm": 0.5458192825317383,
+      "learning_rate": 0.0003971341072278302,
+      "loss": 2.0396,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5795506648326456,
+      "grad_norm": 0.5182467699050903,
+      "learning_rate": 0.00039640741506251457,
+      "loss": 2.0962,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5800091701054562,
+      "grad_norm": 0.5378604531288147,
+      "learning_rate": 0.00039568095146107495,
+      "loss": 1.9702,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5804676753782668,
+      "grad_norm": 0.5289828181266785,
+      "learning_rate": 0.00039495471802636096,
+      "loss": 2.12,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5809261806510775,
+      "grad_norm": 0.5213769674301147,
+      "learning_rate": 0.0003942287163607145,
+      "loss": 1.985,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5813846859238881,
+      "grad_norm": 0.5612925291061401,
+      "learning_rate": 0.0003935029480659658,
+      "loss": 2.0793,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5818431911966988,
+      "grad_norm": 0.5335249900817871,
+      "learning_rate": 0.00039277741474343054,
+      "loss": 2.1316,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5823016964695094,
+      "grad_norm": 0.5264742374420166,
+      "learning_rate": 0.0003920521179939057,
+      "loss": 1.9973,
+      "step": 1270
+    },
+    {
+      "epoch": 0.58276020174232,
+      "grad_norm": 0.504972517490387,
+      "learning_rate": 0.00039132705941766644,
+      "loss": 2.0981,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5832187070151307,
+      "grad_norm": 0.5117753744125366,
+      "learning_rate": 0.0003906022406144624,
+      "loss": 1.9799,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5836772122879413,
+      "grad_norm": 0.5257151126861572,
+      "learning_rate": 0.0003898776631835143,
+      "loss": 2.0882,
+      "step": 1273
+    },
+    {
+      "epoch": 0.584135717560752,
+      "grad_norm": 0.510727047920227,
+      "learning_rate": 0.00038915332872350994,
+      "loss": 1.9276,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5845942228335625,
+      "grad_norm": 0.5261306166648865,
+      "learning_rate": 0.00038842923883260135,
+      "loss": 1.9679,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5850527281063732,
+      "grad_norm": 0.5674136877059937,
+      "learning_rate": 0.00038770539510840093,
+      "loss": 2.066,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5855112333791839,
+      "grad_norm": 0.5441506505012512,
+      "learning_rate": 0.00038698179914797783,
+      "loss": 2.0736,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5859697386519945,
+      "grad_norm": 0.5627047419548035,
+      "learning_rate": 0.0003862584525478545,
+      "loss": 2.0691,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5864282439248051,
+      "grad_norm": 0.5821051597595215,
+      "learning_rate": 0.00038553535690400353,
+      "loss": 2.0943,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5868867491976157,
+      "grad_norm": 0.5146974921226501,
+      "learning_rate": 0.00038481251381184355,
+      "loss": 2.0231,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5873452544704264,
+      "grad_norm": 0.5017138719558716,
+      "learning_rate": 0.00038408992486623584,
+      "loss": 2.0432,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5878037597432371,
+      "grad_norm": 0.5212745070457458,
+      "learning_rate": 0.00038336759166148117,
+      "loss": 2.1016,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5882622650160477,
+      "grad_norm": 0.5024069547653198,
+      "learning_rate": 0.0003826455157913159,
+      "loss": 1.9217,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5887207702888583,
+      "grad_norm": 0.5232961773872375,
+      "learning_rate": 0.00038192369884890886,
+      "loss": 2.0556,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5891792755616689,
+      "grad_norm": 0.5318247675895691,
+      "learning_rate": 0.00038120214242685723,
+      "loss": 1.9995,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5896377808344796,
+      "grad_norm": 0.5133913159370422,
+      "learning_rate": 0.00038048084811718373,
+      "loss": 1.9818,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5900962861072903,
+      "grad_norm": 0.554863691329956,
+      "learning_rate": 0.0003797598175113327,
+      "loss": 2.0255,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5905547913801009,
+      "grad_norm": 0.5218652486801147,
+      "learning_rate": 0.0003790390522001662,
+      "loss": 1.981,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5910132966529115,
+      "grad_norm": 0.5399390459060669,
+      "learning_rate": 0.0003783185537739615,
+      "loss": 2.0796,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5914718019257221,
+      "grad_norm": 0.5524738430976868,
+      "learning_rate": 0.00037759832382240697,
+      "loss": 2.0659,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5919303071985328,
+      "grad_norm": 0.545306921005249,
+      "learning_rate": 0.00037687836393459826,
+      "loss": 2.009,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5923888124713435,
+      "grad_norm": 0.5375888347625732,
+      "learning_rate": 0.0003761586756990354,
+      "loss": 2.0236,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5928473177441541,
+      "grad_norm": 0.5039032101631165,
+      "learning_rate": 0.0003754392607036191,
+      "loss": 2.0648,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5933058230169647,
+      "grad_norm": 0.5127273201942444,
+      "learning_rate": 0.0003747201205356472,
+      "loss": 2.0601,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5937643282897753,
+      "grad_norm": 0.5145084261894226,
+      "learning_rate": 0.0003740012567818111,
+      "loss": 2.1431,
+      "step": 1295
+    },
+    {
+      "epoch": 0.594222833562586,
+      "grad_norm": 0.5352330207824707,
+      "learning_rate": 0.0003732826710281922,
+      "loss": 1.9705,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5946813388353966,
+      "grad_norm": 0.5159630179405212,
+      "learning_rate": 0.0003725643648602588,
+      "loss": 1.9983,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5951398441082072,
+      "grad_norm": 0.5373986959457397,
+      "learning_rate": 0.0003718463398628621,
+      "loss": 2.044,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5955983493810179,
+      "grad_norm": 0.5321218967437744,
+      "learning_rate": 0.0003711285976202331,
+      "loss": 2.0822,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5960568546538285,
+      "grad_norm": 0.5400732159614563,
+      "learning_rate": 0.0003704111397159787,
+      "loss": 1.9897,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5965153599266392,
+      "grad_norm": 0.5253557562828064,
+      "learning_rate": 0.0003696939677330788,
+      "loss": 2.0907,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5969738651994498,
+      "grad_norm": 0.5076909065246582,
+      "learning_rate": 0.00036897708325388213,
+      "loss": 2.003,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5974323704722604,
+      "grad_norm": 0.5343338251113892,
+      "learning_rate": 0.0003682604878601034,
+      "loss": 1.9881,
+      "step": 1303
+    },
+    {
+      "epoch": 0.597890875745071,
+      "grad_norm": 0.5598551034927368,
+      "learning_rate": 0.000367544183132819,
+      "loss": 1.9302,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5983493810178817,
+      "grad_norm": 0.5342590808868408,
+      "learning_rate": 0.0003668281706524645,
+      "loss": 1.923,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5988078862906924,
+      "grad_norm": 0.5430881977081299,
+      "learning_rate": 0.0003661124519988304,
+      "loss": 2.035,
+      "step": 1306
+    },
+    {
+      "epoch": 0.599266391563503,
+      "grad_norm": 0.5175046920776367,
+      "learning_rate": 0.00036539702875105893,
+      "loss": 1.9912,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5997248968363136,
+      "grad_norm": 0.5420910120010376,
+      "learning_rate": 0.0003646819024876406,
+      "loss": 2.0353,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6001834021091242,
+      "grad_norm": 0.5748980641365051,
+      "learning_rate": 0.0003639670747864107,
+      "loss": 2.1139,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6006419073819349,
+      "grad_norm": 0.5321078300476074,
+      "learning_rate": 0.00036325254722454584,
+      "loss": 2.054,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6011004126547456,
+      "grad_norm": 0.6288198828697205,
+      "learning_rate": 0.00036253832137855997,
+      "loss": 2.0184,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6015589179275561,
+      "grad_norm": 0.5236740112304688,
+      "learning_rate": 0.00036182439882430183,
+      "loss": 2.0092,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6020174232003668,
+      "grad_norm": 0.5276572704315186,
+      "learning_rate": 0.00036111078113695096,
+      "loss": 1.9323,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6024759284731774,
+      "grad_norm": 0.5553674697875977,
+      "learning_rate": 0.0003603974698910139,
+      "loss": 1.9628,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6029344337459881,
+      "grad_norm": 0.5391940474510193,
+      "learning_rate": 0.0003596844666603214,
+      "loss": 1.9383,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6033929390187988,
+      "grad_norm": 0.5510250926017761,
+      "learning_rate": 0.0003589717730180245,
+      "loss": 2.0756,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6038514442916093,
+      "grad_norm": 0.5067034959793091,
+      "learning_rate": 0.00035825939053659117,
+      "loss": 2.0738,
+      "step": 1317
+    },
+    {
+      "epoch": 0.60430994956442,
+      "grad_norm": 0.5150887370109558,
+      "learning_rate": 0.00035754732078780273,
+      "loss": 1.9525,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6047684548372306,
+      "grad_norm": 0.5667287707328796,
+      "learning_rate": 0.00035683556534275076,
+      "loss": 2.1108,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6052269601100413,
+      "grad_norm": 0.5381338596343994,
+      "learning_rate": 0.00035612412577183303,
+      "loss": 2.0096,
+      "step": 1320
+    },
+    {
+      "epoch": 0.605685465382852,
+      "grad_norm": 0.5241124629974365,
+      "learning_rate": 0.00035541300364475063,
+      "loss": 1.9509,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6061439706556625,
+      "grad_norm": 0.529976487159729,
+      "learning_rate": 0.0003547022005305043,
+      "loss": 2.0709,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6066024759284732,
+      "grad_norm": 0.524163544178009,
+      "learning_rate": 0.0003539917179973907,
+      "loss": 2.0558,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6070609812012838,
+      "grad_norm": 0.5352849364280701,
+      "learning_rate": 0.00035328155761299917,
+      "loss": 2.0405,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6075194864740945,
+      "grad_norm": 0.522792398929596,
+      "learning_rate": 0.0003525717209442085,
+      "loss": 1.9805,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6079779917469051,
+      "grad_norm": 0.49664103984832764,
+      "learning_rate": 0.00035186220955718306,
+      "loss": 1.8419,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6084364970197157,
+      "grad_norm": 0.530649721622467,
+      "learning_rate": 0.0003511530250173696,
+      "loss": 2.0324,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6088950022925264,
+      "grad_norm": 0.5303981900215149,
+      "learning_rate": 0.00035044416888949364,
+      "loss": 1.979,
+      "step": 1328
+    },
+    {
+      "epoch": 0.609353507565337,
+      "grad_norm": 0.5595840215682983,
+      "learning_rate": 0.0003497356427375562,
+      "loss": 2.0357,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6098120128381477,
+      "grad_norm": 0.5612637996673584,
+      "learning_rate": 0.00034902744812483034,
+      "loss": 2.0668,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6102705181109582,
+      "grad_norm": 0.528218686580658,
+      "learning_rate": 0.00034831958661385714,
+      "loss": 2.0169,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6107290233837689,
+      "grad_norm": 0.516678512096405,
+      "learning_rate": 0.0003476120597664434,
+      "loss": 2.019,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6111875286565795,
+      "grad_norm": 0.534909188747406,
+      "learning_rate": 0.00034690486914365704,
+      "loss": 2.0531,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6116460339293902,
+      "grad_norm": 0.5296339392662048,
+      "learning_rate": 0.00034619801630582435,
+      "loss": 2.0606,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6121045392022009,
+      "grad_norm": 0.5179243087768555,
+      "learning_rate": 0.00034549150281252633,
+      "loss": 1.9896,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6125630444750114,
+      "grad_norm": 0.5275792479515076,
+      "learning_rate": 0.0003447853302225952,
+      "loss": 2.0058,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6130215497478221,
+      "grad_norm": 0.5471377372741699,
+      "learning_rate": 0.00034407950009411126,
+      "loss": 1.9269,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6134800550206327,
+      "grad_norm": 0.5303784012794495,
+      "learning_rate": 0.00034337401398439873,
+      "loss": 1.9718,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6139385602934434,
+      "grad_norm": 0.5493420958518982,
+      "learning_rate": 0.00034266887345002305,
+      "loss": 1.968,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6143970655662541,
+      "grad_norm": 0.5653674006462097,
+      "learning_rate": 0.0003419640800467874,
+      "loss": 2.0594,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6148555708390646,
+      "grad_norm": 0.5535478591918945,
+      "learning_rate": 0.0003412596353297288,
+      "loss": 2.0367,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6153140761118753,
+      "grad_norm": 0.5366288423538208,
+      "learning_rate": 0.00034055554085311493,
+      "loss": 2.0239,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6157725813846859,
+      "grad_norm": 0.5193936228752136,
+      "learning_rate": 0.00033985179817044105,
+      "loss": 2.0687,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6162310866574966,
+      "grad_norm": 0.5410805344581604,
+      "learning_rate": 0.0003391484088344257,
+      "loss": 2.0265,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6166895919303071,
+      "grad_norm": 0.532106339931488,
+      "learning_rate": 0.00033844537439700807,
+      "loss": 2.1277,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6171480972031178,
+      "grad_norm": 0.5267893075942993,
+      "learning_rate": 0.00033774269640934445,
+      "loss": 2.0334,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6176066024759285,
+      "grad_norm": 0.5250502824783325,
+      "learning_rate": 0.0003370403764218045,
+      "loss": 1.9775,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6180651077487391,
+      "grad_norm": 0.5277730226516724,
+      "learning_rate": 0.000336338415983968,
+      "loss": 1.999,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6185236130215498,
+      "grad_norm": 0.503973126411438,
+      "learning_rate": 0.00033563681664462155,
+      "loss": 1.9867,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6189821182943603,
+      "grad_norm": 0.5266078114509583,
+      "learning_rate": 0.000334935579951755,
+      "loss": 2.1402,
+      "step": 1350
+    },
+    {
+      "epoch": 0.619440623567171,
+      "grad_norm": 0.5102262496948242,
+      "learning_rate": 0.0003342347074525578,
+      "loss": 1.9779,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6198991288399817,
+      "grad_norm": 0.5160202980041504,
+      "learning_rate": 0.0003335342006934161,
+      "loss": 2.0684,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6203576341127923,
+      "grad_norm": 0.521024763584137,
+      "learning_rate": 0.00033283406121990914,
+      "loss": 2.0596,
+      "step": 1353
+    },
+    {
+      "epoch": 0.620816139385603,
+      "grad_norm": 0.5350157618522644,
+      "learning_rate": 0.0003321342905768057,
+      "loss": 1.9551,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6212746446584135,
+      "grad_norm": 0.5806898474693298,
+      "learning_rate": 0.00033143489030806086,
+      "loss": 1.9956,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6217331499312242,
+      "grad_norm": 0.5317165851593018,
+      "learning_rate": 0.00033073586195681227,
+      "loss": 2.0067,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6221916552040349,
+      "grad_norm": 0.5820338129997253,
+      "learning_rate": 0.00033003720706537736,
+      "loss": 1.9131,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6226501604768455,
+      "grad_norm": 0.5360593795776367,
+      "learning_rate": 0.0003293389271752492,
+      "loss": 2.0113,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6231086657496562,
+      "grad_norm": 0.5128529667854309,
+      "learning_rate": 0.00032864102382709374,
+      "loss": 2.1169,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6235671710224667,
+      "grad_norm": 0.5368690490722656,
+      "learning_rate": 0.000327943498560746,
+      "loss": 2.0147,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6240256762952774,
+      "grad_norm": 0.5345142483711243,
+      "learning_rate": 0.00032724635291520694,
+      "loss": 2.0971,
+      "step": 1361
+    },
+    {
+      "epoch": 0.624484181568088,
+      "grad_norm": 0.5056533813476562,
+      "learning_rate": 0.00032654958842863967,
+      "loss": 1.9866,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6249426868408987,
+      "grad_norm": 0.4916091561317444,
+      "learning_rate": 0.0003258532066383667,
+      "loss": 1.9397,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6254011921137093,
+      "grad_norm": 0.5304104685783386,
+      "learning_rate": 0.000325157209080866,
+      "loss": 1.9776,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6258596973865199,
+      "grad_norm": 0.5320877432823181,
+      "learning_rate": 0.00032446159729176743,
+      "loss": 2.0484,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6263182026593306,
+      "grad_norm": 0.5311552286148071,
+      "learning_rate": 0.0003237663728058502,
+      "loss": 1.9295,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6267767079321412,
+      "grad_norm": 0.5333547592163086,
+      "learning_rate": 0.0003230715371570389,
+      "loss": 2.0518,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6272352132049519,
+      "grad_norm": 0.5512670874595642,
+      "learning_rate": 0.00032237709187839996,
+      "loss": 1.9371,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6276937184777625,
+      "grad_norm": 0.5293883681297302,
+      "learning_rate": 0.0003216830385021388,
+      "loss": 1.9597,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6281522237505731,
+      "grad_norm": 0.560495138168335,
+      "learning_rate": 0.0003209893785595959,
+      "loss": 2.1266,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6286107290233838,
+      "grad_norm": 0.5196048021316528,
+      "learning_rate": 0.00032029611358124366,
+      "loss": 1.9514,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6290692342961944,
+      "grad_norm": 0.5543765425682068,
+      "learning_rate": 0.00031960324509668336,
+      "loss": 1.9857,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6295277395690051,
+      "grad_norm": 0.5442017316818237,
+      "learning_rate": 0.0003189107746346412,
+      "loss": 2.022,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6299862448418156,
+      "grad_norm": 0.520160436630249,
+      "learning_rate": 0.0003182187037229653,
+      "loss": 1.9903,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6304447501146263,
+      "grad_norm": 0.5124123096466064,
+      "learning_rate": 0.0003175270338886221,
+      "loss": 2.0859,
+      "step": 1375
+    },
+    {
+      "epoch": 0.630903255387437,
+      "grad_norm": 0.5453944206237793,
+      "learning_rate": 0.00031683576665769345,
+      "loss": 2.036,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6313617606602476,
+      "grad_norm": 0.5068944692611694,
+      "learning_rate": 0.0003161449035553724,
+      "loss": 1.8933,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6318202659330582,
+      "grad_norm": 0.5163761377334595,
+      "learning_rate": 0.00031545444610596077,
+      "loss": 2.0643,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6322787712058688,
+      "grad_norm": 0.5104475617408752,
+      "learning_rate": 0.000314764395832865,
+      "loss": 2.0162,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6327372764786795,
+      "grad_norm": 0.5317782163619995,
+      "learning_rate": 0.0003140747542585934,
+      "loss": 1.9548,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6331957817514902,
+      "grad_norm": 0.5453965067863464,
+      "learning_rate": 0.00031338552290475266,
+      "loss": 2.0565,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6336542870243008,
+      "grad_norm": 0.5149431228637695,
+      "learning_rate": 0.00031269670329204396,
+      "loss": 1.9773,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6341127922971114,
+      "grad_norm": 0.5311601161956787,
+      "learning_rate": 0.0003120082969402604,
+      "loss": 2.0781,
+      "step": 1383
+    },
+    {
+      "epoch": 0.634571297569922,
+      "grad_norm": 0.5486698150634766,
+      "learning_rate": 0.00031132030536828314,
+      "loss": 2.0693,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6350298028427327,
+      "grad_norm": 0.5262513160705566,
+      "learning_rate": 0.00031063273009407805,
+      "loss": 2.0074,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6354883081155434,
+      "grad_norm": 0.5388000011444092,
+      "learning_rate": 0.00030994557263469265,
+      "loss": 1.9753,
+      "step": 1386
+    },
+    {
+      "epoch": 0.635946813388354,
+      "grad_norm": 0.5246709585189819,
+      "learning_rate": 0.0003092588345062526,
+      "loss": 2.0266,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6364053186611646,
+      "grad_norm": 0.5381985902786255,
+      "learning_rate": 0.0003085725172239582,
+      "loss": 2.0542,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6368638239339752,
+      "grad_norm": 0.535058319568634,
+      "learning_rate": 0.0003078866223020815,
+      "loss": 2.0079,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6373223292067859,
+      "grad_norm": 0.5505255460739136,
+      "learning_rate": 0.0003072011512539624,
+      "loss": 1.9992,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6377808344795965,
+      "grad_norm": 0.5526959300041199,
+      "learning_rate": 0.00030651610559200574,
+      "loss": 2.0792,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6382393397524071,
+      "grad_norm": 0.49245768785476685,
+      "learning_rate": 0.00030583148682767757,
+      "loss": 1.977,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6386978450252178,
+      "grad_norm": 0.5256136655807495,
+      "learning_rate": 0.00030514729647150243,
+      "loss": 2.0232,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6391563502980284,
+      "grad_norm": 0.5274372696876526,
+      "learning_rate": 0.0003044635360330592,
+      "loss": 1.9756,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6396148555708391,
+      "grad_norm": 0.5110905766487122,
+      "learning_rate": 0.00030378020702097845,
+      "loss": 2.047,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6400733608436497,
+      "grad_norm": 0.5342505574226379,
+      "learning_rate": 0.000303097310942939,
+      "loss": 2.0676,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6405318661164603,
+      "grad_norm": 0.4957382380962372,
+      "learning_rate": 0.0003024148493056641,
+      "loss": 1.9856,
+      "step": 1397
+    },
+    {
+      "epoch": 0.640990371389271,
+      "grad_norm": 0.5313808917999268,
+      "learning_rate": 0.00030173282361491865,
+      "loss": 1.9283,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6414488766620816,
+      "grad_norm": 0.5139783024787903,
+      "learning_rate": 0.0003010512353755057,
+      "loss": 1.9395,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6419073819348923,
+      "grad_norm": 0.4996695816516876,
+      "learning_rate": 0.00030037008609126313,
+      "loss": 1.9824,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6423658872077029,
+      "grad_norm": 0.5511860847473145,
+      "learning_rate": 0.0002996893772650602,
+      "loss": 1.945,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6428243924805135,
+      "grad_norm": 0.5049015879631042,
+      "learning_rate": 0.0002990091103987945,
+      "loss": 2.0843,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6432828977533241,
+      "grad_norm": 0.5585852265357971,
+      "learning_rate": 0.0002983292869933886,
+      "loss": 2.0868,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6437414030261348,
+      "grad_norm": 0.5545941591262817,
+      "learning_rate": 0.0002976499085487862,
+      "loss": 2.0191,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6441999082989455,
+      "grad_norm": 0.5196374654769897,
+      "learning_rate": 0.00029697097656394963,
+      "loss": 2.0168,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6446584135717561,
+      "grad_norm": 0.5293759107589722,
+      "learning_rate": 0.00029629249253685595,
+      "loss": 1.9559,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6451169188445667,
+      "grad_norm": 0.5151011943817139,
+      "learning_rate": 0.00029561445796449416,
+      "loss": 2.0822,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6455754241173773,
+      "grad_norm": 0.5228344202041626,
+      "learning_rate": 0.0002949368743428612,
+      "loss": 1.9521,
+      "step": 1408
+    },
+    {
+      "epoch": 0.646033929390188,
+      "grad_norm": 0.5443601608276367,
+      "learning_rate": 0.0002942597431669593,
+      "loss": 1.9834,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6464924346629987,
+      "grad_norm": 0.5448914766311646,
+      "learning_rate": 0.0002935830659307924,
+      "loss": 2.0215,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6469509399358092,
+      "grad_norm": 0.5161738991737366,
+      "learning_rate": 0.0002929068441273629,
+      "loss": 1.9962,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6474094452086199,
+      "grad_norm": 0.5107429027557373,
+      "learning_rate": 0.0002922310792486681,
+      "loss": 2.0245,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6478679504814305,
+      "grad_norm": 0.5297009944915771,
+      "learning_rate": 0.00029155577278569745,
+      "loss": 2.0,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6483264557542412,
+      "grad_norm": 0.5717682242393494,
+      "learning_rate": 0.00029088092622842895,
+      "loss": 2.0404,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6487849610270519,
+      "grad_norm": 0.536279022693634,
+      "learning_rate": 0.00029020654106582544,
+      "loss": 2.0668,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6492434662998624,
+      "grad_norm": 0.5811988711357117,
+      "learning_rate": 0.0002895326187858326,
+      "loss": 1.9954,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6497019715726731,
+      "grad_norm": 0.5522708892822266,
+      "learning_rate": 0.00028885916087537377,
+      "loss": 2.0269,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6501604768454837,
+      "grad_norm": 0.5218204855918884,
+      "learning_rate": 0.00028818616882034877,
+      "loss": 2.0626,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6506189821182944,
+      "grad_norm": 0.5208260416984558,
+      "learning_rate": 0.0002875136441056286,
+      "loss": 1.9486,
+      "step": 1419
+    },
+    {
+      "epoch": 0.651077487391105,
+      "grad_norm": 0.5203031301498413,
+      "learning_rate": 0.000286841588215054,
+      "loss": 2.0646,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6515359926639156,
+      "grad_norm": 0.5014403462409973,
+      "learning_rate": 0.0002861700026314308,
+      "loss": 2.0286,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6519944979367263,
+      "grad_norm": 0.5365167260169983,
+      "learning_rate": 0.00028549888883652686,
+      "loss": 1.9935,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6524530032095369,
+      "grad_norm": 0.5312653183937073,
+      "learning_rate": 0.00028482824831107,
+      "loss": 2.094,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6529115084823476,
+      "grad_norm": 0.5421769618988037,
+      "learning_rate": 0.000284158082534743,
+      "loss": 1.9614,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6533700137551581,
+      "grad_norm": 0.5185056924819946,
+      "learning_rate": 0.00028348839298618177,
+      "loss": 1.9977,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6538285190279688,
+      "grad_norm": 0.51771479845047,
+      "learning_rate": 0.0002828191811429709,
+      "loss": 2.0015,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6542870243007795,
+      "grad_norm": 0.5269346237182617,
+      "learning_rate": 0.00028215044848164164,
+      "loss": 2.0511,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6547455295735901,
+      "grad_norm": 0.5923379063606262,
+      "learning_rate": 0.00028148219647766747,
+      "loss": 2.0971,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6552040348464008,
+      "grad_norm": 0.570615291595459,
+      "learning_rate": 0.00028081442660546124,
+      "loss": 1.9791,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6556625401192113,
+      "grad_norm": 0.573758602142334,
+      "learning_rate": 0.0002801471403383728,
+      "loss": 1.925,
+      "step": 1430
+    },
+    {
+      "epoch": 0.656121045392022,
+      "grad_norm": 0.5550107955932617,
+      "learning_rate": 0.00027948033914868415,
+      "loss": 2.033,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6565795506648326,
+      "grad_norm": 0.5727185010910034,
+      "learning_rate": 0.00027881402450760775,
+      "loss": 2.0537,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6570380559376433,
+      "grad_norm": 0.5568172931671143,
+      "learning_rate": 0.00027814819788528165,
+      "loss": 2.0828,
+      "step": 1433
+    },
+    {
+      "epoch": 0.657496561210454,
+      "grad_norm": 0.5544043779373169,
+      "learning_rate": 0.00027748286075076836,
+      "loss": 1.9611,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6579550664832645,
+      "grad_norm": 0.524235188961029,
+      "learning_rate": 0.00027681801457204937,
+      "loss": 1.983,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6584135717560752,
+      "grad_norm": 0.5528565645217896,
+      "learning_rate": 0.00027615366081602306,
+      "loss": 2.0589,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6588720770288858,
+      "grad_norm": 0.554866373538971,
+      "learning_rate": 0.0002754898009485021,
+      "loss": 1.9707,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6593305823016965,
+      "grad_norm": 0.5520939826965332,
+      "learning_rate": 0.0002748264364342085,
+      "loss": 1.9708,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6597890875745072,
+      "grad_norm": 0.544072687625885,
+      "learning_rate": 0.00027416356873677204,
+      "loss": 1.9581,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6602475928473177,
+      "grad_norm": 0.5281481146812439,
+      "learning_rate": 0.0002735011993187258,
+      "loss": 1.9672,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6607060981201284,
+      "grad_norm": 0.5461376309394836,
+      "learning_rate": 0.0002728393296415042,
+      "loss": 2.1023,
+      "step": 1441
+    },
+    {
+      "epoch": 0.661164603392939,
+      "grad_norm": 0.5535180568695068,
+      "learning_rate": 0.00027217796116543817,
+      "loss": 2.0051,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6616231086657497,
+      "grad_norm": 0.5356543064117432,
+      "learning_rate": 0.0002715170953497532,
+      "loss": 1.9242,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6620816139385602,
+      "grad_norm": 0.5647934079170227,
+      "learning_rate": 0.00027085673365256614,
+      "loss": 1.9867,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6625401192113709,
+      "grad_norm": 0.5434362888336182,
+      "learning_rate": 0.00027019687753088075,
+      "loss": 2.0606,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6629986244841816,
+      "grad_norm": 0.5759847164154053,
+      "learning_rate": 0.00026953752844058597,
+      "loss": 2.0173,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6634571297569922,
+      "grad_norm": 0.5573680996894836,
+      "learning_rate": 0.0002688786878364516,
+      "loss": 2.0645,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6639156350298029,
+      "grad_norm": 0.5156645178794861,
+      "learning_rate": 0.00026822035717212597,
+      "loss": 2.0189,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6643741403026134,
+      "grad_norm": 0.5184013247489929,
+      "learning_rate": 0.00026756253790013193,
+      "loss": 1.9478,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6648326455754241,
+      "grad_norm": 0.5357776284217834,
+      "learning_rate": 0.0002669052314718641,
+      "loss": 1.967,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6652911508482348,
+      "grad_norm": 0.522457480430603,
+      "learning_rate": 0.0002662484393375855,
+      "loss": 1.9542,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6657496561210454,
+      "grad_norm": 0.5668512582778931,
+      "learning_rate": 0.00026559216294642446,
+      "loss": 2.0083,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6662081613938561,
+      "grad_norm": 0.5374054312705994,
+      "learning_rate": 0.0002649364037463718,
+      "loss": 2.0294,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.5393928289413452,
+      "learning_rate": 0.0002642811631842764,
+      "loss": 2.0111,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6671251719394773,
+      "grad_norm": 0.5165395736694336,
+      "learning_rate": 0.0002636264427058439,
+      "loss": 1.9919,
+      "step": 1455
+    },
+    {
+      "epoch": 0.667583677212288,
+      "grad_norm": 0.5360638499259949,
+      "learning_rate": 0.00026297224375563123,
+      "loss": 1.9544,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6680421824850986,
+      "grad_norm": 0.5196322798728943,
+      "learning_rate": 0.00026231856777704575,
+      "loss": 2.0083,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6685006877579092,
+      "grad_norm": 0.5655248165130615,
+      "learning_rate": 0.00026166541621234026,
+      "loss": 1.9613,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6689591930307198,
+      "grad_norm": 0.5662278532981873,
+      "learning_rate": 0.00026101279050261045,
+      "loss": 1.8926,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6694176983035305,
+      "grad_norm": 0.577921450138092,
+      "learning_rate": 0.00026036069208779247,
+      "loss": 2.0218,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6698762035763411,
+      "grad_norm": 0.5296681523323059,
+      "learning_rate": 0.0002597091224066581,
+      "loss": 1.9875,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6703347088491518,
+      "grad_norm": 0.5702780485153198,
+      "learning_rate": 0.00025905808289681365,
+      "loss": 1.9918,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6707932141219624,
+      "grad_norm": 0.5366291403770447,
+      "learning_rate": 0.0002584075749946946,
+      "loss": 2.0648,
+      "step": 1463
+    },
+    {
+      "epoch": 0.671251719394773,
+      "grad_norm": 0.5491457581520081,
+      "learning_rate": 0.00025775760013556424,
+      "loss": 1.9893,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6717102246675837,
+      "grad_norm": 0.5186455845832825,
+      "learning_rate": 0.0002571081597535095,
+      "loss": 2.0781,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6721687299403943,
+      "grad_norm": 0.540378749370575,
+      "learning_rate": 0.00025645925528143776,
+      "loss": 2.0676,
+      "step": 1466
+    },
+    {
+      "epoch": 0.672627235213205,
+      "grad_norm": 0.5796226263046265,
+      "learning_rate": 0.0002558108881510747,
+      "loss": 2.071,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6730857404860155,
+      "grad_norm": 0.5470061898231506,
+      "learning_rate": 0.00025516305979295963,
+      "loss": 1.9239,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6735442457588262,
+      "grad_norm": 0.5276967883110046,
+      "learning_rate": 0.0002545157716364439,
+      "loss": 2.0509,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6740027510316369,
+      "grad_norm": 0.5172312259674072,
+      "learning_rate": 0.00025386902510968624,
+      "loss": 1.9441,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6744612563044475,
+      "grad_norm": 0.5388919115066528,
+      "learning_rate": 0.00025322282163965095,
+      "loss": 2.0799,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6749197615772582,
+      "grad_norm": 0.5395244359970093,
+      "learning_rate": 0.00025257716265210384,
+      "loss": 2.0109,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6753782668500687,
+      "grad_norm": 0.50630122423172,
+      "learning_rate": 0.0002519320495716091,
+      "loss": 1.9208,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6758367721228794,
+      "grad_norm": 0.5605035424232483,
+      "learning_rate": 0.00025128748382152716,
+      "loss": 2.0866,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6762952773956901,
+      "grad_norm": 0.5117762684822083,
+      "learning_rate": 0.00025064346682401016,
+      "loss": 1.9773,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6767537826685007,
+      "grad_norm": 0.5592853426933289,
+      "learning_rate": 0.0002500000000000001,
+      "loss": 2.0433,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6772122879413113,
+      "grad_norm": 0.564650297164917,
+      "learning_rate": 0.0002493570847692246,
+      "loss": 2.0177,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6776707932141219,
+      "grad_norm": 0.5537674427032471,
+      "learning_rate": 0.00024871472255019424,
+      "loss": 1.9817,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6781292984869326,
+      "grad_norm": 0.5264727473258972,
+      "learning_rate": 0.00024807291476019994,
+      "loss": 2.0112,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6785878037597433,
+      "grad_norm": 0.5421814918518066,
+      "learning_rate": 0.00024743166281530877,
+      "loss": 1.9805,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6790463090325539,
+      "grad_norm": 0.5347735285758972,
+      "learning_rate": 0.000246790968130362,
+      "loss": 2.0246,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6795048143053645,
+      "grad_norm": 0.5414533019065857,
+      "learning_rate": 0.0002461508321189706,
+      "loss": 1.9776,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6799633195781751,
+      "grad_norm": 0.5828797221183777,
+      "learning_rate": 0.00024551125619351385,
+      "loss": 1.9993,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6804218248509858,
+      "grad_norm": 0.5395531058311462,
+      "learning_rate": 0.00024487224176513453,
+      "loss": 2.0012,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6808803301237965,
+      "grad_norm": 0.5179250240325928,
+      "learning_rate": 0.0002442337902437365,
+      "loss": 2.0547,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6813388353966071,
+      "grad_norm": 0.5079845190048218,
+      "learning_rate": 0.0002435959030379824,
+      "loss": 1.9666,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6817973406694177,
+      "grad_norm": 0.5138550996780396,
+      "learning_rate": 0.00024295858155528888,
+      "loss": 1.9903,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6822558459422283,
+      "grad_norm": 0.5129337310791016,
+      "learning_rate": 0.00024232182720182523,
+      "loss": 2.0178,
+      "step": 1488
+    },
+    {
+      "epoch": 0.682714351215039,
+      "grad_norm": 0.5635784864425659,
+      "learning_rate": 0.00024168564138250855,
+      "loss": 2.0427,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6831728564878496,
+      "grad_norm": 0.533216655254364,
+      "learning_rate": 0.00024105002550100246,
+      "loss": 2.0385,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6836313617606602,
+      "grad_norm": 0.5166699886322021,
+      "learning_rate": 0.00024041498095971254,
+      "loss": 2.0464,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6840898670334709,
+      "grad_norm": 0.5407620668411255,
+      "learning_rate": 0.0002397805091597835,
+      "loss": 2.0435,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6845483723062815,
+      "grad_norm": 0.5436777472496033,
+      "learning_rate": 0.0002391466115010973,
+      "loss": 2.0135,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6850068775790922,
+      "grad_norm": 0.5405540466308594,
+      "learning_rate": 0.00023851328938226808,
+      "loss": 2.0166,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6854653828519028,
+      "grad_norm": 0.569582462310791,
+      "learning_rate": 0.00023788054420064109,
+      "loss": 1.9347,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6859238881247134,
+      "grad_norm": 0.5252484679222107,
+      "learning_rate": 0.00023724837735228773,
+      "loss": 1.9832,
+      "step": 1496
+    },
+    {
+      "epoch": 0.686382393397524,
+      "grad_norm": 0.5349242091178894,
+      "learning_rate": 0.00023661679023200422,
+      "loss": 2.0474,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6868408986703347,
+      "grad_norm": 0.5523650050163269,
+      "learning_rate": 0.00023598578423330714,
+      "loss": 1.9543,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6872994039431454,
+      "grad_norm": 0.5316759347915649,
+      "learning_rate": 0.00023535536074843083,
+      "loss": 1.9779,
+      "step": 1499
+    },
+    {
+      "epoch": 0.687757909215956,
+      "grad_norm": 0.5272276401519775,
+      "learning_rate": 0.00023472552116832502,
+      "loss": 2.0724,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6882164144887666,
+      "grad_norm": 0.5480548143386841,
+      "learning_rate": 0.0002340962668826503,
+      "loss": 2.0248,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6886749197615772,
+      "grad_norm": 0.5266236066818237,
+      "learning_rate": 0.00023346759927977663,
+      "loss": 2.0656,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6891334250343879,
+      "grad_norm": 0.5382277369499207,
+      "learning_rate": 0.0002328395197467789,
+      "loss": 1.9995,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6895919303071986,
+      "grad_norm": 0.5341281890869141,
+      "learning_rate": 0.00023221202966943515,
+      "loss": 2.034,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6900504355800092,
+      "grad_norm": 0.5351428389549255,
+      "learning_rate": 0.0002315851304322223,
+      "loss": 2.0242,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6905089408528198,
+      "grad_norm": 0.5410450100898743,
+      "learning_rate": 0.0002309588234183137,
+      "loss": 2.0408,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6909674461256304,
+      "grad_norm": 0.5219683051109314,
+      "learning_rate": 0.00023033311000957653,
+      "loss": 1.976,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6914259513984411,
+      "grad_norm": 0.5376242399215698,
+      "learning_rate": 0.00022970799158656758,
+      "loss": 1.9844,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6918844566712518,
+      "grad_norm": 0.544317364692688,
+      "learning_rate": 0.0002290834695285316,
+      "loss": 2.0158,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6923429619440623,
+      "grad_norm": 0.5282340049743652,
+      "learning_rate": 0.00022845954521339678,
+      "loss": 2.0759,
+      "step": 1510
+    },
+    {
+      "epoch": 0.692801467216873,
+      "grad_norm": 0.5466699600219727,
+      "learning_rate": 0.0002278362200177732,
+      "loss": 2.0325,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6932599724896836,
+      "grad_norm": 0.5225258469581604,
+      "learning_rate": 0.00022721349531694852,
+      "loss": 1.9802,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6937184777624943,
+      "grad_norm": 0.5675475001335144,
+      "learning_rate": 0.0002265913724848855,
+      "loss": 1.9823,
+      "step": 1513
+    },
+    {
+      "epoch": 0.694176983035305,
+      "grad_norm": 0.5370016098022461,
+      "learning_rate": 0.00022596985289421946,
+      "loss": 2.0113,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6946354883081155,
+      "grad_norm": 0.5250517725944519,
+      "learning_rate": 0.00022534893791625405,
+      "loss": 1.9676,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6950939935809262,
+      "grad_norm": 0.5523369908332825,
+      "learning_rate": 0.00022472862892095968,
+      "loss": 1.9972,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6955524988537368,
+      "grad_norm": 0.558120846748352,
+      "learning_rate": 0.00022410892727696896,
+      "loss": 2.0682,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6960110041265475,
+      "grad_norm": 0.5363593101501465,
+      "learning_rate": 0.0002234898343515751,
+      "loss": 2.0279,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6964695093993581,
+      "grad_norm": 0.5090158581733704,
+      "learning_rate": 0.00022287135151072792,
+      "loss": 1.953,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6969280146721687,
+      "grad_norm": 0.5171800255775452,
+      "learning_rate": 0.00022225348011903096,
+      "loss": 1.9648,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6973865199449794,
+      "grad_norm": 0.5427493453025818,
+      "learning_rate": 0.0002216362215397393,
+      "loss": 1.9367,
+      "step": 1521
+    },
+    {
+      "epoch": 0.69784502521779,
+      "grad_norm": 0.5350388288497925,
+      "learning_rate": 0.00022101957713475522,
+      "loss": 1.9665,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6983035304906007,
+      "grad_norm": 0.5283173322677612,
+      "learning_rate": 0.00022040354826462666,
+      "loss": 1.9827,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6987620357634112,
+      "grad_norm": 0.5425599217414856,
+      "learning_rate": 0.0002197881362885426,
+      "loss": 1.9818,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6992205410362219,
+      "grad_norm": 0.5350459814071655,
+      "learning_rate": 0.0002191733425643318,
+      "loss": 1.913,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6996790463090325,
+      "grad_norm": 0.5171528458595276,
+      "learning_rate": 0.00021855916844845826,
+      "loss": 1.9742,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7001375515818432,
+      "grad_norm": 0.5030922293663025,
+      "learning_rate": 0.00021794561529601898,
+      "loss": 1.9388,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7005960568546539,
+      "grad_norm": 0.537511944770813,
+      "learning_rate": 0.00021733268446074138,
+      "loss": 1.9839,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7010545621274644,
+      "grad_norm": 0.5231597423553467,
+      "learning_rate": 0.00021672037729497917,
+      "loss": 1.9539,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7015130674002751,
+      "grad_norm": 0.5410900115966797,
+      "learning_rate": 0.0002161086951497106,
+      "loss": 2.0751,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7019715726730857,
+      "grad_norm": 0.5727490186691284,
+      "learning_rate": 0.00021549763937453442,
+      "loss": 2.0564,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7024300779458964,
+      "grad_norm": 0.5589496493339539,
+      "learning_rate": 0.00021488721131766736,
+      "loss": 2.0274,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7028885832187071,
+      "grad_norm": 0.5308485627174377,
+      "learning_rate": 0.00021427741232594183,
+      "loss": 1.9579,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7033470884915176,
+      "grad_norm": 0.5481025576591492,
+      "learning_rate": 0.0002136682437448013,
+      "loss": 1.9968,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7038055937643283,
+      "grad_norm": 0.5186814665794373,
+      "learning_rate": 0.0002130597069182994,
+      "loss": 1.9299,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7042640990371389,
+      "grad_norm": 0.5166908502578735,
+      "learning_rate": 0.0002124518031890948,
+      "loss": 1.9949,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7047226043099496,
+      "grad_norm": 0.4986916780471802,
+      "learning_rate": 0.0002118445338984502,
+      "loss": 2.0253,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7051811095827603,
+      "grad_norm": 0.5607078075408936,
+      "learning_rate": 0.00021123790038622808,
+      "loss": 2.0322,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7056396148555708,
+      "grad_norm": 0.5386072993278503,
+      "learning_rate": 0.0002106319039908879,
+      "loss": 2.1537,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7060981201283815,
+      "grad_norm": 0.49631157517433167,
+      "learning_rate": 0.00021002654604948412,
+      "loss": 1.9998,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7065566254011921,
+      "grad_norm": 0.526479184627533,
+      "learning_rate": 0.00020942182789766172,
+      "loss": 1.9875,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7070151306740028,
+      "grad_norm": 0.5390573143959045,
+      "learning_rate": 0.00020881775086965492,
+      "loss": 1.9963,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7074736359468133,
+      "grad_norm": 0.518315851688385,
+      "learning_rate": 0.00020821431629828246,
+      "loss": 1.9556,
+      "step": 1543
+    },
+    {
+      "epoch": 0.707932141219624,
+      "grad_norm": 0.555034339427948,
+      "learning_rate": 0.00020761152551494643,
+      "loss": 1.998,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7083906464924347,
+      "grad_norm": 0.5053515434265137,
+      "learning_rate": 0.00020700937984962798,
+      "loss": 1.9655,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7088491517652453,
+      "grad_norm": 0.5342771410942078,
+      "learning_rate": 0.0002064078806308848,
+      "loss": 2.1012,
+      "step": 1546
+    },
+    {
+      "epoch": 0.709307657038056,
+      "grad_norm": 0.5490493774414062,
+      "learning_rate": 0.00020580702918584882,
+      "loss": 2.0868,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7097661623108665,
+      "grad_norm": 0.5208424925804138,
+      "learning_rate": 0.000205206826840222,
+      "loss": 1.913,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7102246675836772,
+      "grad_norm": 0.5568975806236267,
+      "learning_rate": 0.0002046072749182751,
+      "loss": 2.0913,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7106831728564879,
+      "grad_norm": 0.547223687171936,
+      "learning_rate": 0.00020400837474284273,
+      "loss": 1.9913,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7111416781292985,
+      "grad_norm": 0.5454710721969604,
+      "learning_rate": 0.0002034101276353224,
+      "loss": 2.0222,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7116001834021092,
+      "grad_norm": 0.5384700298309326,
+      "learning_rate": 0.00020281253491567027,
+      "loss": 2.0524,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7120586886749197,
+      "grad_norm": 0.5073917508125305,
+      "learning_rate": 0.0002022155979023984,
+      "loss": 1.9932,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7125171939477304,
+      "grad_norm": 0.5165788531303406,
+      "learning_rate": 0.000201619317912573,
+      "loss": 2.0132,
+      "step": 1554
+    },
+    {
+      "epoch": 0.712975699220541,
+      "grad_norm": 0.531200110912323,
+      "learning_rate": 0.00020102369626180962,
+      "loss": 1.9661,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7134342044933517,
+      "grad_norm": 0.529605507850647,
+      "learning_rate": 0.0002004287342642721,
+      "loss": 1.9689,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7138927097661623,
+      "grad_norm": 0.5542986989021301,
+      "learning_rate": 0.00019983443323266824,
+      "loss": 2.1177,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7143512150389729,
+      "grad_norm": 0.528816819190979,
+      "learning_rate": 0.00019924079447824805,
+      "loss": 2.0061,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7148097203117836,
+      "grad_norm": 0.5439432859420776,
+      "learning_rate": 0.00019864781931079977,
+      "loss": 1.9544,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7152682255845942,
+      "grad_norm": 0.5292602181434631,
+      "learning_rate": 0.00019805550903864773,
+      "loss": 2.0231,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7157267308574049,
+      "grad_norm": 0.5466258525848389,
+      "learning_rate": 0.00019746386496864948,
+      "loss": 1.9714,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7161852361302155,
+      "grad_norm": 0.548058807849884,
+      "learning_rate": 0.00019687288840619226,
+      "loss": 1.9838,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7166437414030261,
+      "grad_norm": 0.5433715581893921,
+      "learning_rate": 0.0001962825806551911,
+      "loss": 1.94,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7171022466758368,
+      "grad_norm": 0.5429866313934326,
+      "learning_rate": 0.0001956929430180846,
+      "loss": 1.9438,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7175607519486474,
+      "grad_norm": 0.5164733529090881,
+      "learning_rate": 0.00019510397679583374,
+      "loss": 2.0041,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7180192572214581,
+      "grad_norm": 0.5603983998298645,
+      "learning_rate": 0.0001945156832879174,
+      "loss": 1.938,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7184777624942686,
+      "grad_norm": 0.5343077778816223,
+      "learning_rate": 0.00019392806379233036,
+      "loss": 2.0432,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7189362677670793,
+      "grad_norm": 0.5447087287902832,
+      "learning_rate": 0.00019334111960558065,
+      "loss": 1.8714,
+      "step": 1568
+    },
+    {
+      "epoch": 0.71939477303989,
+      "grad_norm": 0.5458582639694214,
+      "learning_rate": 0.00019275485202268573,
+      "loss": 2.0792,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7198532783127006,
+      "grad_norm": 0.5358990430831909,
+      "learning_rate": 0.00019216926233717085,
+      "loss": 1.905,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7203117835855112,
+      "grad_norm": 0.5501385927200317,
+      "learning_rate": 0.00019158435184106498,
+      "loss": 2.0498,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7207702888583218,
+      "grad_norm": 0.5453598499298096,
+      "learning_rate": 0.00019100012182489905,
+      "loss": 1.9966,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7212287941311325,
+      "grad_norm": 0.5118822455406189,
+      "learning_rate": 0.00019041657357770226,
+      "loss": 1.9955,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7216872994039432,
+      "grad_norm": 0.49914059042930603,
+      "learning_rate": 0.00018983370838699943,
+      "loss": 1.9899,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7221458046767538,
+      "grad_norm": 0.5381051898002625,
+      "learning_rate": 0.00018925152753880892,
+      "loss": 1.9405,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7226043099495644,
+      "grad_norm": 0.5078611373901367,
+      "learning_rate": 0.00018867003231763847,
+      "loss": 2.0283,
+      "step": 1576
+    },
+    {
+      "epoch": 0.723062815222375,
+      "grad_norm": 0.5207589864730835,
+      "learning_rate": 0.00018808922400648375,
+      "loss": 2.0042,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7235213204951857,
+      "grad_norm": 0.5199658870697021,
+      "learning_rate": 0.00018750910388682428,
+      "loss": 1.9783,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7239798257679964,
+      "grad_norm": 0.5393915176391602,
+      "learning_rate": 0.00018692967323862125,
+      "loss": 1.9734,
+      "step": 1579
+    },
+    {
+      "epoch": 0.724438331040807,
+      "grad_norm": 0.5441501140594482,
+      "learning_rate": 0.00018635093334031517,
+      "loss": 1.9585,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7248968363136176,
+      "grad_norm": 0.5385583639144897,
+      "learning_rate": 0.00018577288546882165,
+      "loss": 1.91,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7253553415864282,
+      "grad_norm": 0.5472796559333801,
+      "learning_rate": 0.00018519553089953023,
+      "loss": 2.0039,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7258138468592389,
+      "grad_norm": 0.5197535157203674,
+      "learning_rate": 0.0001846188709063001,
+      "loss": 2.0473,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7262723521320495,
+      "grad_norm": 0.5530803203582764,
+      "learning_rate": 0.00018404290676145857,
+      "loss": 2.0409,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7267308574048602,
+      "grad_norm": 0.5449008345603943,
+      "learning_rate": 0.00018346763973579722,
+      "loss": 1.9462,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7271893626776708,
+      "grad_norm": 0.5619111657142639,
+      "learning_rate": 0.00018289307109856939,
+      "loss": 1.9962,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7276478679504814,
+      "grad_norm": 0.5299919247627258,
+      "learning_rate": 0.0001823192021174882,
+      "loss": 1.9823,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7281063732232921,
+      "grad_norm": 0.5335145592689514,
+      "learning_rate": 0.0001817460340587223,
+      "loss": 2.0298,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7285648784961027,
+      "grad_norm": 0.5164056420326233,
+      "learning_rate": 0.00018117356818689445,
+      "loss": 1.9445,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7290233837689133,
+      "grad_norm": 0.531017005443573,
+      "learning_rate": 0.00018060180576507756,
+      "loss": 2.0028,
+      "step": 1590
+    },
+    {
+      "epoch": 0.729481889041724,
+      "grad_norm": 0.5519788265228271,
+      "learning_rate": 0.00018003074805479313,
+      "loss": 2.0844,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7299403943145346,
+      "grad_norm": 0.5700712203979492,
+      "learning_rate": 0.00017946039631600724,
+      "loss": 1.9817,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7303988995873453,
+      "grad_norm": 0.5224321484565735,
+      "learning_rate": 0.00017889075180712837,
+      "loss": 2.0228,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7308574048601559,
+      "grad_norm": 0.5358400344848633,
+      "learning_rate": 0.00017832181578500512,
+      "loss": 1.9889,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7313159101329665,
+      "grad_norm": 0.5245378017425537,
+      "learning_rate": 0.0001777535895049221,
+      "loss": 1.91,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7317744154057771,
+      "grad_norm": 0.5320974588394165,
+      "learning_rate": 0.0001771860742205988,
+      "loss": 2.0098,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7322329206785878,
+      "grad_norm": 0.5469915866851807,
+      "learning_rate": 0.00017661927118418525,
+      "loss": 1.9631,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7326914259513985,
+      "grad_norm": 0.5484734773635864,
+      "learning_rate": 0.00017605318164626066,
+      "loss": 1.9824,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7331499312242091,
+      "grad_norm": 0.5294070243835449,
+      "learning_rate": 0.00017548780685582949,
+      "loss": 2.0757,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7336084364970197,
+      "grad_norm": 0.5723766684532166,
+      "learning_rate": 0.00017492314806031922,
+      "loss": 2.0999,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7340669417698303,
+      "grad_norm": 0.5182992219924927,
+      "learning_rate": 0.00017435920650557806,
+      "loss": 1.919,
+      "step": 1601
+    },
+    {
+      "epoch": 0.734525447042641,
+      "grad_norm": 0.5128610134124756,
+      "learning_rate": 0.00017379598343587112,
+      "loss": 2.0203,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7349839523154517,
+      "grad_norm": 0.5214248299598694,
+      "learning_rate": 0.00017323348009387878,
+      "loss": 2.0285,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7354424575882622,
+      "grad_norm": 0.5153412818908691,
+      "learning_rate": 0.0001726716977206929,
+      "loss": 1.9776,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7359009628610729,
+      "grad_norm": 0.5317767858505249,
+      "learning_rate": 0.00017211063755581525,
+      "loss": 1.9496,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7363594681338835,
+      "grad_norm": 0.5347232818603516,
+      "learning_rate": 0.0001715503008371536,
+      "loss": 1.9998,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7368179734066942,
+      "grad_norm": 0.5121638178825378,
+      "learning_rate": 0.0001709906888010196,
+      "loss": 2.0004,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7372764786795049,
+      "grad_norm": 0.5310835242271423,
+      "learning_rate": 0.00017043180268212638,
+      "loss": 1.9865,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7377349839523154,
+      "grad_norm": 0.5444746017456055,
+      "learning_rate": 0.00016987364371358481,
+      "loss": 1.9458,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7381934892251261,
+      "grad_norm": 0.5703752040863037,
+      "learning_rate": 0.00016931621312690214,
+      "loss": 2.0352,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7386519944979367,
+      "grad_norm": 0.5737705230712891,
+      "learning_rate": 0.00016875951215197777,
+      "loss": 2.0166,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7391104997707474,
+      "grad_norm": 0.5496467351913452,
+      "learning_rate": 0.00016820354201710214,
+      "loss": 1.9609,
+      "step": 1612
+    },
+    {
+      "epoch": 0.739569005043558,
+      "grad_norm": 0.5437312126159668,
+      "learning_rate": 0.00016764830394895203,
+      "loss": 2.0228,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7400275103163686,
+      "grad_norm": 0.5542490482330322,
+      "learning_rate": 0.00016709379917259027,
+      "loss": 2.0092,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7404860155891793,
+      "grad_norm": 0.5283743739128113,
+      "learning_rate": 0.00016654002891146091,
+      "loss": 1.9976,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7409445208619899,
+      "grad_norm": 0.5338504910469055,
+      "learning_rate": 0.00016598699438738764,
+      "loss": 1.96,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7414030261348006,
+      "grad_norm": 0.5790355205535889,
+      "learning_rate": 0.00016543469682057105,
+      "loss": 2.043,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7418615314076112,
+      "grad_norm": 0.5363684892654419,
+      "learning_rate": 0.00016488313742958526,
+      "loss": 2.0231,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7423200366804218,
+      "grad_norm": 0.5220614075660706,
+      "learning_rate": 0.00016433231743137646,
+      "loss": 2.0404,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7427785419532325,
+      "grad_norm": 0.5090878009796143,
+      "learning_rate": 0.0001637822380412584,
+      "loss": 1.9818,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7432370472260431,
+      "grad_norm": 0.5535461902618408,
+      "learning_rate": 0.00016323290047291195,
+      "loss": 1.9309,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7436955524988538,
+      "grad_norm": 0.521217405796051,
+      "learning_rate": 0.0001626843059383803,
+      "loss": 2.0815,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7441540577716643,
+      "grad_norm": 0.544882595539093,
+      "learning_rate": 0.00016213645564806752,
+      "loss": 2.0522,
+      "step": 1623
+    },
+    {
+      "epoch": 0.744612563044475,
+      "grad_norm": 0.5379012823104858,
+      "learning_rate": 0.0001615893508107359,
+      "loss": 1.878,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7450710683172856,
+      "grad_norm": 0.5120853185653687,
+      "learning_rate": 0.00016104299263350252,
+      "loss": 2.0368,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7455295735900963,
+      "grad_norm": 0.5202006101608276,
+      "learning_rate": 0.00016049738232183758,
+      "loss": 1.9684,
+      "step": 1626
+    },
+    {
+      "epoch": 0.745988078862907,
+      "grad_norm": 0.5289000868797302,
+      "learning_rate": 0.0001599525210795606,
+      "loss": 1.8962,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7464465841357175,
+      "grad_norm": 0.5122154355049133,
+      "learning_rate": 0.00015940841010883889,
+      "loss": 2.0272,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7469050894085282,
+      "grad_norm": 0.5651560425758362,
+      "learning_rate": 0.00015886505061018413,
+      "loss": 2.0147,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7473635946813388,
+      "grad_norm": 0.54373699426651,
+      "learning_rate": 0.0001583224437824498,
+      "loss": 1.967,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7478220999541495,
+      "grad_norm": 0.5371326804161072,
+      "learning_rate": 0.0001577805908228293,
+      "loss": 2.0661,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7482806052269602,
+      "grad_norm": 0.5702128410339355,
+      "learning_rate": 0.00015723949292685191,
+      "loss": 1.9731,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7487391104997707,
+      "grad_norm": 0.5438613891601562,
+      "learning_rate": 0.0001566991512883818,
+      "loss": 1.9193,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7491976157725814,
+      "grad_norm": 0.5430235266685486,
+      "learning_rate": 0.00015615956709961378,
+      "loss": 2.0958,
+      "step": 1634
+    },
+    {
+      "epoch": 0.749656121045392,
+      "grad_norm": 0.5889632701873779,
+      "learning_rate": 0.00015562074155107215,
+      "loss": 1.9496,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7501146263182027,
+      "grad_norm": 0.5804908871650696,
+      "learning_rate": 0.0001550826758316068,
+      "loss": 2.0121,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7505731315910132,
+      "grad_norm": 0.5805303454399109,
+      "learning_rate": 0.00015454537112839122,
+      "loss": 2.0456,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7510316368638239,
+      "grad_norm": 0.5228239297866821,
+      "learning_rate": 0.00015400882862692033,
+      "loss": 2.0228,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7514901421366346,
+      "grad_norm": 0.5319299697875977,
+      "learning_rate": 0.00015347304951100665,
+      "loss": 1.9952,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7519486474094452,
+      "grad_norm": 0.520284116268158,
+      "learning_rate": 0.00015293803496277907,
+      "loss": 2.0545,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7524071526822559,
+      "grad_norm": 0.5642551779747009,
+      "learning_rate": 0.00015240378616267886,
+      "loss": 1.9989,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7528656579550664,
+      "grad_norm": 0.5528300404548645,
+      "learning_rate": 0.00015187030428945843,
+      "loss": 1.915,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7533241632278771,
+      "grad_norm": 0.5486469268798828,
+      "learning_rate": 0.0001513375905201776,
+      "loss": 2.0263,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7537826685006878,
+      "grad_norm": 0.5280976295471191,
+      "learning_rate": 0.00015080564603020142,
+      "loss": 1.9837,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7542411737734984,
+      "grad_norm": 0.5559554100036621,
+      "learning_rate": 0.0001502744719931982,
+      "loss": 1.9182,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7546996790463091,
+      "grad_norm": 0.5170323848724365,
+      "learning_rate": 0.00014974406958113558,
+      "loss": 2.0149,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7551581843191196,
+      "grad_norm": 0.5329339504241943,
+      "learning_rate": 0.00014921443996427947,
+      "loss": 1.992,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7556166895919303,
+      "grad_norm": 0.5411702990531921,
+      "learning_rate": 0.0001486855843111901,
+      "loss": 2.0561,
+      "step": 1648
+    },
+    {
+      "epoch": 0.756075194864741,
+      "grad_norm": 0.516733705997467,
+      "learning_rate": 0.0001481575037887201,
+      "loss": 2.0434,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7565337001375516,
+      "grad_norm": 0.5125020146369934,
+      "learning_rate": 0.00014763019956201253,
+      "loss": 2.0411,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7569922054103623,
+      "grad_norm": 0.5202275514602661,
+      "learning_rate": 0.0001471036727944966,
+      "loss": 1.9549,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7574507106831728,
+      "grad_norm": 0.5001369714736938,
+      "learning_rate": 0.0001465779246478872,
+      "loss": 1.9015,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7579092159559835,
+      "grad_norm": 0.5139358043670654,
+      "learning_rate": 0.00014605295628218045,
+      "loss": 1.9252,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7583677212287941,
+      "grad_norm": 0.5397683382034302,
+      "learning_rate": 0.0001455287688556527,
+      "loss": 1.9546,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7588262265016048,
+      "grad_norm": 0.5189513564109802,
+      "learning_rate": 0.00014500536352485673,
+      "loss": 1.9538,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7592847317744154,
+      "grad_norm": 0.5130738019943237,
+      "learning_rate": 0.00014448274144461965,
+      "loss": 1.9342,
+      "step": 1656
+    },
+    {
+      "epoch": 0.759743237047226,
+      "grad_norm": 0.5209282040596008,
+      "learning_rate": 0.00014396090376804112,
+      "loss": 2.0355,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7602017423200367,
+      "grad_norm": 0.5216522812843323,
+      "learning_rate": 0.00014343985164648926,
+      "loss": 1.9472,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7606602475928473,
+      "grad_norm": 0.5561063289642334,
+      "learning_rate": 0.00014291958622959973,
+      "loss": 2.1497,
+      "step": 1659
+    },
+    {
+      "epoch": 0.761118752865658,
+      "grad_norm": 0.580458402633667,
+      "learning_rate": 0.00014240010866527176,
+      "loss": 2.0198,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7615772581384685,
+      "grad_norm": 0.522511899471283,
+      "learning_rate": 0.00014188142009966686,
+      "loss": 2.0223,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7620357634112792,
+      "grad_norm": 0.5592499375343323,
+      "learning_rate": 0.0001413635216772053,
+      "loss": 1.9474,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7624942686840899,
+      "grad_norm": 0.5315657258033752,
+      "learning_rate": 0.000140846414540564,
+      "loss": 2.0314,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7629527739569005,
+      "grad_norm": 0.5378697514533997,
+      "learning_rate": 0.00014033009983067452,
+      "loss": 1.9589,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7634112792297112,
+      "grad_norm": 0.5324126482009888,
+      "learning_rate": 0.00013981457868671927,
+      "loss": 1.9634,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7638697845025217,
+      "grad_norm": 0.5258311629295349,
+      "learning_rate": 0.0001392998522461305,
+      "loss": 1.9981,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7643282897753324,
+      "grad_norm": 0.565293550491333,
+      "learning_rate": 0.00013878592164458635,
+      "loss": 2.0303,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7647867950481431,
+      "grad_norm": 0.5282187461853027,
+      "learning_rate": 0.00013827278801600978,
+      "loss": 1.9705,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7652453003209537,
+      "grad_norm": 0.5262969136238098,
+      "learning_rate": 0.0001377604524925647,
+      "loss": 1.978,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7657038055937643,
+      "grad_norm": 0.5400656461715698,
+      "learning_rate": 0.00013724891620465424,
+      "loss": 1.9783,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7661623108665749,
+      "grad_norm": 0.5201109051704407,
+      "learning_rate": 0.0001367381802809185,
+      "loss": 1.9628,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7666208161393856,
+      "grad_norm": 0.5359255075454712,
+      "learning_rate": 0.00013622824584823113,
+      "loss": 1.9564,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7670793214121963,
+      "grad_norm": 0.5274782180786133,
+      "learning_rate": 0.00013571911403169795,
+      "loss": 2.0183,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7675378266850069,
+      "grad_norm": 0.5776581168174744,
+      "learning_rate": 0.0001352107859546533,
+      "loss": 1.9868,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7679963319578175,
+      "grad_norm": 0.5260794758796692,
+      "learning_rate": 0.00013470326273865886,
+      "loss": 2.0033,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7684548372306281,
+      "grad_norm": 0.5149418711662292,
+      "learning_rate": 0.00013419654550349985,
+      "loss": 1.9408,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7689133425034388,
+      "grad_norm": 0.5315396189689636,
+      "learning_rate": 0.00013369063536718346,
+      "loss": 1.9215,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7693718477762495,
+      "grad_norm": 0.5422107577323914,
+      "learning_rate": 0.00013318553344593632,
+      "loss": 1.9579,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7698303530490601,
+      "grad_norm": 0.5203865170478821,
+      "learning_rate": 0.00013268124085420136,
+      "loss": 1.9378,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7702888583218707,
+      "grad_norm": 0.5393579602241516,
+      "learning_rate": 0.0001321777587046364,
+      "loss": 1.9755,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7707473635946813,
+      "grad_norm": 0.5563924312591553,
+      "learning_rate": 0.00013167508810811059,
+      "loss": 1.9864,
+      "step": 1681
+    },
+    {
+      "epoch": 0.771205868867492,
+      "grad_norm": 0.5501232147216797,
+      "learning_rate": 0.0001311732301737029,
+      "loss": 2.0173,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7716643741403026,
+      "grad_norm": 0.5072229504585266,
+      "learning_rate": 0.0001306721860086991,
+      "loss": 1.9847,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7721228794131133,
+      "grad_norm": 0.5172702670097351,
+      "learning_rate": 0.00013017195671858928,
+      "loss": 1.9203,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7725813846859239,
+      "grad_norm": 0.5172668099403381,
+      "learning_rate": 0.0001296725434070661,
+      "loss": 1.9358,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7730398899587345,
+      "grad_norm": 0.5118347406387329,
+      "learning_rate": 0.00012917394717602121,
+      "loss": 1.9543,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7734983952315452,
+      "grad_norm": 0.5347263813018799,
+      "learning_rate": 0.00012867616912554426,
+      "loss": 2.065,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7739569005043558,
+      "grad_norm": 0.5374776124954224,
+      "learning_rate": 0.00012817921035391882,
+      "loss": 1.9281,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7744154057771664,
+      "grad_norm": 0.5431856513023376,
+      "learning_rate": 0.00012768307195762168,
+      "loss": 1.9795,
+      "step": 1689
+    },
+    {
+      "epoch": 0.774873911049977,
+      "grad_norm": 0.5269105434417725,
+      "learning_rate": 0.00012718775503131908,
+      "loss": 1.9893,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7753324163227877,
+      "grad_norm": 0.5374506115913391,
+      "learning_rate": 0.0001266932606678646,
+      "loss": 1.9474,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7757909215955984,
+      "grad_norm": 0.5754181146621704,
+      "learning_rate": 0.00012619958995829756,
+      "loss": 1.9515,
+      "step": 1692
+    },
+    {
+      "epoch": 0.776249426868409,
+      "grad_norm": 0.5318741202354431,
+      "learning_rate": 0.0001257067439918394,
+      "loss": 1.9543,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7767079321412196,
+      "grad_norm": 0.525360107421875,
+      "learning_rate": 0.00012521472385589234,
+      "loss": 2.0143,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7771664374140302,
+      "grad_norm": 0.5531063675880432,
+      "learning_rate": 0.00012472353063603626,
+      "loss": 1.9695,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7776249426868409,
+      "grad_norm": 0.5372726321220398,
+      "learning_rate": 0.0001242331654160263,
+      "loss": 1.9628,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7780834479596516,
+      "grad_norm": 0.5618669986724854,
+      "learning_rate": 0.0001237436292777914,
+      "loss": 2.0029,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7785419532324622,
+      "grad_norm": 0.5451949834823608,
+      "learning_rate": 0.00012325492330143061,
+      "loss": 2.0045,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7790004585052728,
+      "grad_norm": 0.5526834726333618,
+      "learning_rate": 0.00012276704856521175,
+      "loss": 2.0266,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7794589637780834,
+      "grad_norm": 0.5471678376197815,
+      "learning_rate": 0.00012228000614556816,
+      "loss": 1.9333,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7799174690508941,
+      "grad_norm": 0.5532666444778442,
+      "learning_rate": 0.00012179379711709738,
+      "loss": 1.8677,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7803759743237048,
+      "grad_norm": 0.5294289588928223,
+      "learning_rate": 0.0001213084225525577,
+      "loss": 2.115,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7808344795965153,
+      "grad_norm": 0.5254940390586853,
+      "learning_rate": 0.00012082388352286627,
+      "loss": 1.9737,
+      "step": 1703
+    },
+    {
+      "epoch": 0.781292984869326,
+      "grad_norm": 0.5349456071853638,
+      "learning_rate": 0.00012034018109709716,
+      "loss": 2.0059,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7817514901421366,
+      "grad_norm": 0.529731035232544,
+      "learning_rate": 0.00011985731634247809,
+      "loss": 2.0394,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7822099954149473,
+      "grad_norm": 0.5313467383384705,
+      "learning_rate": 0.00011937529032438904,
+      "loss": 1.9596,
+      "step": 1706
+    },
+    {
+      "epoch": 0.782668500687758,
+      "grad_norm": 0.5250717997550964,
+      "learning_rate": 0.00011889410410635887,
+      "loss": 1.9543,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7831270059605685,
+      "grad_norm": 0.5252338647842407,
+      "learning_rate": 0.0001184137587500641,
+      "loss": 2.0013,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7835855112333792,
+      "grad_norm": 0.5565682053565979,
+      "learning_rate": 0.00011793425531532564,
+      "loss": 1.9332,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7840440165061898,
+      "grad_norm": 0.5106825828552246,
+      "learning_rate": 0.00011745559486010671,
+      "loss": 2.0792,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7845025217790005,
+      "grad_norm": 0.5336278676986694,
+      "learning_rate": 0.00011697777844051105,
+      "loss": 1.9551,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7849610270518111,
+      "grad_norm": 0.5357179641723633,
+      "learning_rate": 0.00011650080711077964,
+      "loss": 1.962,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7854195323246217,
+      "grad_norm": 0.5396364331245422,
+      "learning_rate": 0.00011602468192328936,
+      "loss": 1.9864,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7858780375974324,
+      "grad_norm": 0.5142610669136047,
+      "learning_rate": 0.00011554940392854973,
+      "loss": 1.9809,
+      "step": 1714
+    },
+    {
+      "epoch": 0.786336542870243,
+      "grad_norm": 0.576097309589386,
+      "learning_rate": 0.00011507497417520146,
+      "loss": 2.041,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7867950481430537,
+      "grad_norm": 0.5536897778511047,
+      "learning_rate": 0.00011460139371001339,
+      "loss": 2.0075,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7872535534158643,
+      "grad_norm": 0.5152952075004578,
+      "learning_rate": 0.00011412866357788049,
+      "loss": 1.9799,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7877120586886749,
+      "grad_norm": 0.5536704659461975,
+      "learning_rate": 0.00011365678482182207,
+      "loss": 2.0332,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7881705639614855,
+      "grad_norm": 0.5463110208511353,
+      "learning_rate": 0.0001131857584829783,
+      "loss": 2.0018,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7886290692342962,
+      "grad_norm": 0.5497841835021973,
+      "learning_rate": 0.0001127155856006093,
+      "loss": 1.9491,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7890875745071069,
+      "grad_norm": 0.5130734443664551,
+      "learning_rate": 0.00011224626721209141,
+      "loss": 1.9944,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7895460797799174,
+      "grad_norm": 0.5367828607559204,
+      "learning_rate": 0.0001117778043529164,
+      "loss": 2.0066,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7900045850527281,
+      "grad_norm": 0.5295432209968567,
+      "learning_rate": 0.0001113101980566879,
+      "loss": 1.9966,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7904630903255387,
+      "grad_norm": 0.5429772138595581,
+      "learning_rate": 0.00011084344935511958,
+      "loss": 1.9294,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7909215955983494,
+      "grad_norm": 0.5457319617271423,
+      "learning_rate": 0.00011037755927803345,
+      "loss": 2.0205,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7913801008711601,
+      "grad_norm": 0.5125073194503784,
+      "learning_rate": 0.00010991252885335651,
+      "loss": 1.9913,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7918386061439706,
+      "grad_norm": 0.5400171875953674,
+      "learning_rate": 0.00010944835910711958,
+      "loss": 1.9469,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7922971114167813,
+      "grad_norm": 0.5607929229736328,
+      "learning_rate": 0.00010898505106345396,
+      "loss": 2.0015,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7927556166895919,
+      "grad_norm": 0.519460916519165,
+      "learning_rate": 0.00010852260574459022,
+      "loss": 1.9378,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7932141219624026,
+      "grad_norm": 0.5683660507202148,
+      "learning_rate": 0.00010806102417085512,
+      "loss": 2.0177,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7936726272352133,
+      "grad_norm": 0.5506643056869507,
+      "learning_rate": 0.00010760030736066951,
+      "loss": 2.034,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7941311325080238,
+      "grad_norm": 0.5728784203529358,
+      "learning_rate": 0.00010714045633054687,
+      "loss": 1.9833,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7945896377808345,
+      "grad_norm": 0.57149738073349,
+      "learning_rate": 0.00010668147209508971,
+      "loss": 1.9478,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7950481430536451,
+      "grad_norm": 0.5502626299858093,
+      "learning_rate": 0.00010622335566698877,
+      "loss": 2.0791,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7955066483264558,
+      "grad_norm": 0.5020186305046082,
+      "learning_rate": 0.00010576610805701942,
+      "loss": 2.0166,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7959651535992663,
+      "grad_norm": 0.5536559820175171,
+      "learning_rate": 0.00010530973027404073,
+      "loss": 1.9728,
+      "step": 1736
+    },
+    {
+      "epoch": 0.796423658872077,
+      "grad_norm": 0.5346456170082092,
+      "learning_rate": 0.00010485422332499212,
+      "loss": 2.0219,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7968821641448877,
+      "grad_norm": 0.5301395058631897,
+      "learning_rate": 0.00010439958821489165,
+      "loss": 1.992,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7973406694176983,
+      "grad_norm": 0.5200710296630859,
+      "learning_rate": 0.00010394582594683428,
+      "loss": 2.0597,
+      "step": 1739
+    },
+    {
+      "epoch": 0.797799174690509,
+      "grad_norm": 0.5282193422317505,
+      "learning_rate": 0.0001034929375219884,
+      "loss": 1.9486,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7982576799633195,
+      "grad_norm": 0.5301898717880249,
+      "learning_rate": 0.00010304092393959514,
+      "loss": 1.9093,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7987161852361302,
+      "grad_norm": 0.490508109331131,
+      "learning_rate": 0.00010258978619696468,
+      "loss": 1.9726,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7991746905089409,
+      "grad_norm": 0.5290476083755493,
+      "learning_rate": 0.00010213952528947551,
+      "loss": 2.039,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7996331957817515,
+      "grad_norm": 0.5471696853637695,
+      "learning_rate": 0.00010169014221057089,
+      "loss": 1.9591,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8000917010545622,
+      "grad_norm": 0.5246416926383972,
+      "learning_rate": 0.00010124163795175734,
+      "loss": 1.9725,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8005502063273727,
+      "grad_norm": 0.5293288826942444,
+      "learning_rate": 0.00010079401350260287,
+      "loss": 2.0206,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8010087116001834,
+      "grad_norm": 0.5461154580116272,
+      "learning_rate": 0.00010034726985073362,
+      "loss": 2.0243,
+      "step": 1747
+    },
+    {
+      "epoch": 0.801467216872994,
+      "grad_norm": 0.5469599366188049,
+      "learning_rate": 9.9901407981833e-05,
+      "loss": 2.0344,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8019257221458047,
+      "grad_norm": 0.5402487516403198,
+      "learning_rate": 9.94564288796384e-05,
+      "loss": 1.9387,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8023842274186153,
+      "grad_norm": 0.521140456199646,
+      "learning_rate": 9.901233352593953e-05,
+      "loss": 2.0269,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8028427326914259,
+      "grad_norm": 0.5508837699890137,
+      "learning_rate": 9.856912290057668e-05,
+      "loss": 1.9379,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8033012379642366,
+      "grad_norm": 0.5221036076545715,
+      "learning_rate": 9.812679798143748e-05,
+      "loss": 1.994,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8037597432370472,
+      "grad_norm": 0.5611392855644226,
+      "learning_rate": 9.768535974445586e-05,
+      "loss": 2.0253,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8042182485098579,
+      "grad_norm": 0.5493707656860352,
+      "learning_rate": 9.724480916360906e-05,
+      "loss": 1.9695,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8046767537826685,
+      "grad_norm": 0.5372111797332764,
+      "learning_rate": 9.68051472109162e-05,
+      "loss": 2.0706,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8051352590554791,
+      "grad_norm": 0.535719096660614,
+      "learning_rate": 9.636637485643529e-05,
+      "loss": 1.9344,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8055937643282898,
+      "grad_norm": 0.5333956480026245,
+      "learning_rate": 9.592849306826174e-05,
+      "loss": 1.9512,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8060522696011004,
+      "grad_norm": 0.557956874370575,
+      "learning_rate": 9.549150281252633e-05,
+      "loss": 2.0757,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8065107748739111,
+      "grad_norm": 0.5516635179519653,
+      "learning_rate": 9.505540505339223e-05,
+      "loss": 1.9155,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8069692801467216,
+      "grad_norm": 0.5757764577865601,
+      "learning_rate": 9.4620200753054e-05,
+      "loss": 2.0112,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8074277854195323,
+      "grad_norm": 0.5421778559684753,
+      "learning_rate": 9.418589087173441e-05,
+      "loss": 1.9739,
+      "step": 1761
+    },
+    {
+      "epoch": 0.807886290692343,
+      "grad_norm": 0.5514983534812927,
+      "learning_rate": 9.375247636768325e-05,
+      "loss": 1.9279,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8083447959651536,
+      "grad_norm": 0.5474739670753479,
+      "learning_rate": 9.331995819717443e-05,
+      "loss": 2.0103,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8088033012379643,
+      "grad_norm": 0.5566092729568481,
+      "learning_rate": 9.288833731450419e-05,
+      "loss": 1.9713,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8092618065107748,
+      "grad_norm": 0.5322392582893372,
+      "learning_rate": 9.245761467198948e-05,
+      "loss": 1.9675,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8097203117835855,
+      "grad_norm": 0.5494855642318726,
+      "learning_rate": 9.20277912199648e-05,
+      "loss": 1.8914,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8101788170563962,
+      "grad_norm": 0.5335338711738586,
+      "learning_rate": 9.159886790678123e-05,
+      "loss": 1.9866,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8106373223292068,
+      "grad_norm": 0.5289465188980103,
+      "learning_rate": 9.11708456788033e-05,
+      "loss": 1.9529,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8110958276020174,
+      "grad_norm": 0.4994891285896301,
+      "learning_rate": 9.074372548040793e-05,
+      "loss": 1.8935,
+      "step": 1769
+    },
+    {
+      "epoch": 0.811554332874828,
+      "grad_norm": 0.5275200605392456,
+      "learning_rate": 9.031750825398145e-05,
+      "loss": 1.9818,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8120128381476387,
+      "grad_norm": 0.5147053003311157,
+      "learning_rate": 8.98921949399179e-05,
+      "loss": 2.0011,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8124713434204494,
+      "grad_norm": 0.5394434332847595,
+      "learning_rate": 8.94677864766173e-05,
+      "loss": 1.9505,
+      "step": 1772
+    },
+    {
+      "epoch": 0.81292984869326,
+      "grad_norm": 0.5584526658058167,
+      "learning_rate": 8.904428380048269e-05,
+      "loss": 1.9946,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8133883539660706,
+      "grad_norm": 0.546439528465271,
+      "learning_rate": 8.862168784591929e-05,
+      "loss": 1.9903,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8138468592388812,
+      "grad_norm": 0.5576072335243225,
+      "learning_rate": 8.819999954533115e-05,
+      "loss": 1.9727,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8143053645116919,
+      "grad_norm": 0.5307096242904663,
+      "learning_rate": 8.777921982911996e-05,
+      "loss": 1.9076,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8147638697845025,
+      "grad_norm": 0.562481164932251,
+      "learning_rate": 8.735934962568253e-05,
+      "loss": 1.9814,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8152223750573132,
+      "grad_norm": 0.5551573634147644,
+      "learning_rate": 8.694038986140945e-05,
+      "loss": 1.9951,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8156808803301238,
+      "grad_norm": 0.5528231263160706,
+      "learning_rate": 8.652234146068206e-05,
+      "loss": 1.9678,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8161393856029344,
+      "grad_norm": 0.539352297782898,
+      "learning_rate": 8.610520534587086e-05,
+      "loss": 1.9271,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8165978908757451,
+      "grad_norm": 0.5565946102142334,
+      "learning_rate": 8.568898243733397e-05,
+      "loss": 1.9519,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8170563961485557,
+      "grad_norm": 0.5632263422012329,
+      "learning_rate": 8.527367365341409e-05,
+      "loss": 2.0326,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8175149014213663,
+      "grad_norm": 0.5997861623764038,
+      "learning_rate": 8.485927991043757e-05,
+      "loss": 2.0513,
+      "step": 1783
+    },
+    {
+      "epoch": 0.817973406694177,
+      "grad_norm": 0.5521172285079956,
+      "learning_rate": 8.444580212271125e-05,
+      "loss": 2.0319,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8184319119669876,
+      "grad_norm": 0.5249801874160767,
+      "learning_rate": 8.403324120252159e-05,
+      "loss": 2.0336,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8188904172397983,
+      "grad_norm": 0.5486734509468079,
+      "learning_rate": 8.362159806013175e-05,
+      "loss": 1.9527,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8193489225126089,
+      "grad_norm": 0.5334566831588745,
+      "learning_rate": 8.321087360377988e-05,
+      "loss": 1.9339,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8198074277854195,
+      "grad_norm": 0.5652158260345459,
+      "learning_rate": 8.280106873967752e-05,
+      "loss": 2.0698,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8202659330582301,
+      "grad_norm": 0.5256199240684509,
+      "learning_rate": 8.239218437200679e-05,
+      "loss": 1.9813,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8207244383310408,
+      "grad_norm": 0.542815089225769,
+      "learning_rate": 8.198422140291939e-05,
+      "loss": 1.9602,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8211829436038515,
+      "grad_norm": 0.5322232246398926,
+      "learning_rate": 8.157718073253351e-05,
+      "loss": 2.0303,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8216414488766621,
+      "grad_norm": 0.5741298794746399,
+      "learning_rate": 8.117106325893287e-05,
+      "loss": 1.9481,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8220999541494727,
+      "grad_norm": 0.5451644659042358,
+      "learning_rate": 8.076586987816404e-05,
+      "loss": 1.9801,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8225584594222833,
+      "grad_norm": 0.5533276200294495,
+      "learning_rate": 8.036160148423449e-05,
+      "loss": 1.9492,
+      "step": 1794
+    },
+    {
+      "epoch": 0.823016964695094,
+      "grad_norm": 0.5420897603034973,
+      "learning_rate": 7.995825896911141e-05,
+      "loss": 1.9993,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8234754699679047,
+      "grad_norm": 0.5311235785484314,
+      "learning_rate": 7.955584322271853e-05,
+      "loss": 1.9943,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8239339752407153,
+      "grad_norm": 0.5092657208442688,
+      "learning_rate": 7.915435513293523e-05,
+      "loss": 1.9782,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8243924805135259,
+      "grad_norm": 0.5569383502006531,
+      "learning_rate": 7.875379558559387e-05,
+      "loss": 1.8598,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8248509857863365,
+      "grad_norm": 0.5370429754257202,
+      "learning_rate": 7.835416546447838e-05,
+      "loss": 1.9619,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8253094910591472,
+      "grad_norm": 0.537312924861908,
+      "learning_rate": 7.795546565132167e-05,
+      "loss": 1.9604,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8257679963319579,
+      "grad_norm": 0.5612266063690186,
+      "learning_rate": 7.755769702580412e-05,
+      "loss": 1.9887,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8262265016047684,
+      "grad_norm": 0.5336949825286865,
+      "learning_rate": 7.716086046555193e-05,
+      "loss": 1.9745,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8266850068775791,
+      "grad_norm": 0.5238338112831116,
+      "learning_rate": 7.676495684613432e-05,
+      "loss": 2.0014,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8271435121503897,
+      "grad_norm": 0.5277724266052246,
+      "learning_rate": 7.636998704106252e-05,
+      "loss": 1.9015,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8276020174232004,
+      "grad_norm": 0.5227574110031128,
+      "learning_rate": 7.597595192178702e-05,
+      "loss": 1.9079,
+      "step": 1805
+    },
+    {
+      "epoch": 0.828060522696011,
+      "grad_norm": 0.536152720451355,
+      "learning_rate": 7.558285235769646e-05,
+      "loss": 1.9497,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8285190279688216,
+      "grad_norm": 0.5545056462287903,
+      "learning_rate": 7.519068921611494e-05,
+      "loss": 2.0614,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8289775332416323,
+      "grad_norm": 0.5590724349021912,
+      "learning_rate": 7.479946336230047e-05,
+      "loss": 1.9615,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8294360385144429,
+      "grad_norm": 0.5411460399627686,
+      "learning_rate": 7.440917565944349e-05,
+      "loss": 2.0324,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8298945437872536,
+      "grad_norm": 0.5358926057815552,
+      "learning_rate": 7.4019826968664e-05,
+      "loss": 1.9355,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8303530490600642,
+      "grad_norm": 0.5284121036529541,
+      "learning_rate": 7.363141814901053e-05,
+      "loss": 1.9441,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8308115543328748,
+      "grad_norm": 0.499912291765213,
+      "learning_rate": 7.32439500574577e-05,
+      "loss": 2.0042,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8312700596056855,
+      "grad_norm": 0.514972984790802,
+      "learning_rate": 7.285742354890473e-05,
+      "loss": 1.9435,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8317285648784961,
+      "grad_norm": 0.5210375785827637,
+      "learning_rate": 7.247183947617325e-05,
+      "loss": 1.9639,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8321870701513068,
+      "grad_norm": 0.521888256072998,
+      "learning_rate": 7.20871986900053e-05,
+      "loss": 1.9557,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8326455754241173,
+      "grad_norm": 0.5275126695632935,
+      "learning_rate": 7.170350203906218e-05,
+      "loss": 1.9132,
+      "step": 1816
+    },
+    {
+      "epoch": 0.833104080696928,
+      "grad_norm": 0.5196903944015503,
+      "learning_rate": 7.132075036992158e-05,
+      "loss": 1.9546,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8335625859697386,
+      "grad_norm": 0.5589911937713623,
+      "learning_rate": 7.093894452707666e-05,
+      "loss": 1.9096,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8340210912425493,
+      "grad_norm": 0.5172871351242065,
+      "learning_rate": 7.055808535293334e-05,
+      "loss": 1.8859,
+      "step": 1819
+    },
+    {
+      "epoch": 0.83447959651536,
+      "grad_norm": 0.5169002413749695,
+      "learning_rate": 7.017817368780888e-05,
+      "loss": 2.0115,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8349381017881705,
+      "grad_norm": 0.4915698170661926,
+      "learning_rate": 6.979921036993042e-05,
+      "loss": 1.9626,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8353966070609812,
+      "grad_norm": 0.5383006930351257,
+      "learning_rate": 6.942119623543202e-05,
+      "loss": 2.0676,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8358551123337918,
+      "grad_norm": 0.5509033203125,
+      "learning_rate": 6.904413211835414e-05,
+      "loss": 1.9901,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8363136176066025,
+      "grad_norm": 0.5662304759025574,
+      "learning_rate": 6.866801885064056e-05,
+      "loss": 1.9982,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8367721228794132,
+      "grad_norm": 0.5635647773742676,
+      "learning_rate": 6.829285726213769e-05,
+      "loss": 1.9575,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8372306281522237,
+      "grad_norm": 0.5774449110031128,
+      "learning_rate": 6.79186481805918e-05,
+      "loss": 2.0385,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8376891334250344,
+      "grad_norm": 0.545552670955658,
+      "learning_rate": 6.754539243164754e-05,
+      "loss": 1.9654,
+      "step": 1827
+    },
+    {
+      "epoch": 0.838147638697845,
+      "grad_norm": 0.537452757358551,
+      "learning_rate": 6.717309083884654e-05,
+      "loss": 1.9853,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8386061439706557,
+      "grad_norm": 0.5465030074119568,
+      "learning_rate": 6.680174422362468e-05,
+      "loss": 1.8846,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8390646492434664,
+      "grad_norm": 0.5350578427314758,
+      "learning_rate": 6.643135340531136e-05,
+      "loss": 1.9938,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8395231545162769,
+      "grad_norm": 0.5530080795288086,
+      "learning_rate": 6.606191920112664e-05,
+      "loss": 2.0759,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8399816597890876,
+      "grad_norm": 0.5762832164764404,
+      "learning_rate": 6.569344242618036e-05,
+      "loss": 1.9052,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8404401650618982,
+      "grad_norm": 0.543536901473999,
+      "learning_rate": 6.532592389346958e-05,
+      "loss": 2.0736,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8408986703347089,
+      "grad_norm": 0.5407480597496033,
+      "learning_rate": 6.495936441387713e-05,
+      "loss": 1.8906,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8413571756075194,
+      "grad_norm": 0.5501353740692139,
+      "learning_rate": 6.459376479617013e-05,
+      "loss": 1.9256,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8418156808803301,
+      "grad_norm": 0.5304622650146484,
+      "learning_rate": 6.422912584699752e-05,
+      "loss": 2.0016,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8422741861531408,
+      "grad_norm": 0.5131492018699646,
+      "learning_rate": 6.386544837088904e-05,
+      "loss": 1.9289,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8427326914259514,
+      "grad_norm": 0.5457130074501038,
+      "learning_rate": 6.350273317025251e-05,
+      "loss": 1.9055,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8431911966987621,
+      "grad_norm": 0.5345990657806396,
+      "learning_rate": 6.314098104537324e-05,
+      "loss": 1.9823,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8436497019715726,
+      "grad_norm": 0.5281217694282532,
+      "learning_rate": 6.278019279441122e-05,
+      "loss": 1.9868,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8441082072443833,
+      "grad_norm": 0.544705331325531,
+      "learning_rate": 6.242036921339972e-05,
+      "loss": 1.9776,
+      "step": 1841
+    },
+    {
+      "epoch": 0.844566712517194,
+      "grad_norm": 0.561013400554657,
+      "learning_rate": 6.206151109624402e-05,
+      "loss": 1.9384,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8450252177900046,
+      "grad_norm": 0.5346787571907043,
+      "learning_rate": 6.170361923471868e-05,
+      "loss": 1.9345,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8454837230628153,
+      "grad_norm": 0.5075468420982361,
+      "learning_rate": 6.134669441846691e-05,
+      "loss": 1.9535,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8459422283356258,
+      "grad_norm": 0.5177434086799622,
+      "learning_rate": 6.099073743499772e-05,
+      "loss": 1.992,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8464007336084365,
+      "grad_norm": 0.5656879544258118,
+      "learning_rate": 6.063574906968511e-05,
+      "loss": 1.9414,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8468592388812471,
+      "grad_norm": 0.5543491244316101,
+      "learning_rate": 6.028173010576582e-05,
+      "loss": 1.9299,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8473177441540578,
+      "grad_norm": 0.52487713098526,
+      "learning_rate": 5.9928681324337544e-05,
+      "loss": 1.9996,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8477762494268684,
+      "grad_norm": 0.5276233553886414,
+      "learning_rate": 5.957660350435773e-05,
+      "loss": 1.9346,
+      "step": 1849
+    },
+    {
+      "epoch": 0.848234754699679,
+      "grad_norm": 0.5655543208122253,
+      "learning_rate": 5.922549742264122e-05,
+      "loss": 2.0104,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8486932599724897,
+      "grad_norm": 0.5383417010307312,
+      "learning_rate": 5.8875363853859166e-05,
+      "loss": 1.9363,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8491517652453003,
+      "grad_norm": 0.5331195592880249,
+      "learning_rate": 5.852620357053651e-05,
+      "loss": 2.001,
+      "step": 1852
+    },
+    {
+      "epoch": 0.849610270518111,
+      "grad_norm": 0.5856356620788574,
+      "learning_rate": 5.8178017343051336e-05,
+      "loss": 1.8963,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8500687757909215,
+      "grad_norm": 0.5179303884506226,
+      "learning_rate": 5.783080593963219e-05,
+      "loss": 1.9752,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8505272810637322,
+      "grad_norm": 0.5461146831512451,
+      "learning_rate": 5.748457012635683e-05,
+      "loss": 1.9856,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8509857863365429,
+      "grad_norm": 0.51792973279953,
+      "learning_rate": 5.713931066715078e-05,
+      "loss": 1.864,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8514442916093535,
+      "grad_norm": 0.4893184304237366,
+      "learning_rate": 5.679502832378497e-05,
+      "loss": 2.0441,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8519027968821642,
+      "grad_norm": 0.5561646819114685,
+      "learning_rate": 5.645172385587482e-05,
+      "loss": 1.9227,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8523613021549747,
+      "grad_norm": 0.5172981023788452,
+      "learning_rate": 5.6109398020877834e-05,
+      "loss": 1.8957,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8528198074277854,
+      "grad_norm": 0.5820702314376831,
+      "learning_rate": 5.576805157409265e-05,
+      "loss": 1.9286,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8532783127005961,
+      "grad_norm": 0.5396410226821899,
+      "learning_rate": 5.542768526865677e-05,
+      "loss": 1.9129,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8537368179734067,
+      "grad_norm": 0.5712278485298157,
+      "learning_rate": 5.508829985554509e-05,
+      "loss": 2.0521,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8541953232462174,
+      "grad_norm": 0.5084553956985474,
+      "learning_rate": 5.474989608356856e-05,
+      "loss": 2.0379,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8546538285190279,
+      "grad_norm": 0.5268993377685547,
+      "learning_rate": 5.441247469937194e-05,
+      "loss": 1.9443,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8551123337918386,
+      "grad_norm": 0.5187596082687378,
+      "learning_rate": 5.407603644743286e-05,
+      "loss": 1.944,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8555708390646493,
+      "grad_norm": 0.5271098613739014,
+      "learning_rate": 5.374058207005944e-05,
+      "loss": 1.886,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8560293443374599,
+      "grad_norm": 0.5565198659896851,
+      "learning_rate": 5.3406112307389066e-05,
+      "loss": 1.9729,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8564878496102705,
+      "grad_norm": 0.5604119300842285,
+      "learning_rate": 5.3072627897386926e-05,
+      "loss": 1.998,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8569463548830811,
+      "grad_norm": 0.5427196621894836,
+      "learning_rate": 5.27401295758439e-05,
+      "loss": 1.9684,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8574048601558918,
+      "grad_norm": 0.5372345447540283,
+      "learning_rate": 5.2408618076375315e-05,
+      "loss": 2.0567,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8578633654287025,
+      "grad_norm": 0.5308073163032532,
+      "learning_rate": 5.207809413041914e-05,
+      "loss": 2.0173,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8583218707015131,
+      "grad_norm": 0.5105649828910828,
+      "learning_rate": 5.174855846723459e-05,
+      "loss": 1.961,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8587803759743237,
+      "grad_norm": 0.5482373833656311,
+      "learning_rate": 5.1420011813900104e-05,
+      "loss": 1.9921,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8592388812471343,
+      "grad_norm": 0.5412172079086304,
+      "learning_rate": 5.109245489531211e-05,
+      "loss": 1.9778,
+      "step": 1874
+    },
+    {
+      "epoch": 0.859697386519945,
+      "grad_norm": 0.5178170800209045,
+      "learning_rate": 5.0765888434183446e-05,
+      "loss": 1.8592,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8601558917927556,
+      "grad_norm": 0.53026282787323,
+      "learning_rate": 5.0440313151041364e-05,
+      "loss": 1.9876,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8606143970655663,
+      "grad_norm": 0.536217451095581,
+      "learning_rate": 5.011572976422657e-05,
+      "loss": 1.8821,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8610729023383769,
+      "grad_norm": 0.5488402843475342,
+      "learning_rate": 4.9792138989890825e-05,
+      "loss": 1.9963,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8615314076111875,
+      "grad_norm": 0.5519595146179199,
+      "learning_rate": 4.9469541541996234e-05,
+      "loss": 1.9349,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8619899128839982,
+      "grad_norm": 0.5345006585121155,
+      "learning_rate": 4.914793813231305e-05,
+      "loss": 2.0171,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8624484181568088,
+      "grad_norm": 0.5429441332817078,
+      "learning_rate": 4.882732947041818e-05,
+      "loss": 2.0165,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8629069234296194,
+      "grad_norm": 0.5263943672180176,
+      "learning_rate": 4.850771626369416e-05,
+      "loss": 1.9837,
+      "step": 1882
+    },
+    {
+      "epoch": 0.86336542870243,
+      "grad_norm": 0.5503644347190857,
+      "learning_rate": 4.818909921732662e-05,
+      "loss": 1.9669,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8638239339752407,
+      "grad_norm": 0.5472528338432312,
+      "learning_rate": 4.787147903430383e-05,
+      "loss": 1.9642,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8642824392480514,
+      "grad_norm": 0.5496426224708557,
+      "learning_rate": 4.755485641541424e-05,
+      "loss": 1.9327,
+      "step": 1885
+    },
+    {
+      "epoch": 0.864740944520862,
+      "grad_norm": 0.560735285282135,
+      "learning_rate": 4.723923205924557e-05,
+      "loss": 1.9111,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8651994497936726,
+      "grad_norm": 0.5405287742614746,
+      "learning_rate": 4.6924606662182736e-05,
+      "loss": 1.979,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8656579550664832,
+      "grad_norm": 0.5257533192634583,
+      "learning_rate": 4.6610980918406596e-05,
+      "loss": 1.9482,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8661164603392939,
+      "grad_norm": 0.5385393500328064,
+      "learning_rate": 4.629835551989276e-05,
+      "loss": 2.0025,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8665749656121046,
+      "grad_norm": 0.5197558403015137,
+      "learning_rate": 4.5986731156409224e-05,
+      "loss": 1.9935,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8670334708849152,
+      "grad_norm": 0.5551241040229797,
+      "learning_rate": 4.567610851551568e-05,
+      "loss": 2.0741,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8674919761577258,
+      "grad_norm": 0.535892903804779,
+      "learning_rate": 4.536648828256146e-05,
+      "loss": 1.99,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8679504814305364,
+      "grad_norm": 0.5309919714927673,
+      "learning_rate": 4.505787114068433e-05,
+      "loss": 1.9521,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8684089867033471,
+      "grad_norm": 0.5399839282035828,
+      "learning_rate": 4.4750257770808764e-05,
+      "loss": 2.0287,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8688674919761578,
+      "grad_norm": 0.5477420091629028,
+      "learning_rate": 4.444364885164448e-05,
+      "loss": 1.9115,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8693259972489684,
+      "grad_norm": 0.5563164949417114,
+      "learning_rate": 4.413804505968533e-05,
+      "loss": 2.0055,
+      "step": 1896
+    },
+    {
+      "epoch": 0.869784502521779,
+      "grad_norm": 0.5260708928108215,
+      "learning_rate": 4.3833447069206944e-05,
+      "loss": 2.035,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8702430077945896,
+      "grad_norm": 0.5468084216117859,
+      "learning_rate": 4.352985555226635e-05,
+      "loss": 2.0415,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8707015130674003,
+      "grad_norm": 0.5307693481445312,
+      "learning_rate": 4.322727117869951e-05,
+      "loss": 1.9453,
+      "step": 1899
+    },
+    {
+      "epoch": 0.871160018340211,
+      "grad_norm": 0.5248668193817139,
+      "learning_rate": 4.29256946161205e-05,
+      "loss": 1.9411,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8716185236130215,
+      "grad_norm": 0.5037166476249695,
+      "learning_rate": 4.262512652991968e-05,
+      "loss": 1.8972,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8720770288858322,
+      "grad_norm": 0.516660213470459,
+      "learning_rate": 4.2325567583262113e-05,
+      "loss": 1.9422,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8725355341586428,
+      "grad_norm": 0.5509219169616699,
+      "learning_rate": 4.2027018437086895e-05,
+      "loss": 1.9208,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8729940394314535,
+      "grad_norm": 0.5533407926559448,
+      "learning_rate": 4.172947975010449e-05,
+      "loss": 1.9002,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8734525447042641,
+      "grad_norm": 0.5449783802032471,
+      "learning_rate": 4.143295217879645e-05,
+      "loss": 2.0814,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8739110499770747,
+      "grad_norm": 0.5420216917991638,
+      "learning_rate": 4.113743637741296e-05,
+      "loss": 1.9908,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8743695552498854,
+      "grad_norm": 0.5403022766113281,
+      "learning_rate": 4.084293299797226e-05,
+      "loss": 1.9962,
+      "step": 1907
+    },
+    {
+      "epoch": 0.874828060522696,
+      "grad_norm": 0.5390098690986633,
+      "learning_rate": 4.054944269025862e-05,
+      "loss": 1.9167,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8752865657955067,
+      "grad_norm": 0.5296696424484253,
+      "learning_rate": 4.025696610182095e-05,
+      "loss": 1.8838,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8757450710683173,
+      "grad_norm": 0.5415107607841492,
+      "learning_rate": 3.996550387797187e-05,
+      "loss": 1.9199,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8762035763411279,
+      "grad_norm": 0.5594709515571594,
+      "learning_rate": 3.9675056661785556e-05,
+      "loss": 1.892,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8766620816139385,
+      "grad_norm": 0.5167399644851685,
+      "learning_rate": 3.9385625094097154e-05,
+      "loss": 1.9936,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8771205868867492,
+      "grad_norm": 0.5205084085464478,
+      "learning_rate": 3.909720981350034e-05,
+      "loss": 1.8857,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8775790921595599,
+      "grad_norm": 0.550877571105957,
+      "learning_rate": 3.880981145634704e-05,
+      "loss": 2.0609,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8780375974323704,
+      "grad_norm": 0.5458064079284668,
+      "learning_rate": 3.852343065674507e-05,
+      "loss": 2.0005,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8784961027051811,
+      "grad_norm": 0.5400051474571228,
+      "learning_rate": 3.8238068046557276e-05,
+      "loss": 1.9247,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8789546079779917,
+      "grad_norm": 0.5541701912879944,
+      "learning_rate": 3.795372425540006e-05,
+      "loss": 1.9243,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8794131132508024,
+      "grad_norm": 0.5226370692253113,
+      "learning_rate": 3.76703999106418e-05,
+      "loss": 1.9274,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8798716185236131,
+      "grad_norm": 0.5125284790992737,
+      "learning_rate": 3.7388095637401754e-05,
+      "loss": 1.9036,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8803301237964236,
+      "grad_norm": 0.5070911049842834,
+      "learning_rate": 3.7106812058548376e-05,
+      "loss": 1.9436,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8807886290692343,
+      "grad_norm": 0.5562089681625366,
+      "learning_rate": 3.682654979469807e-05,
+      "loss": 1.9225,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8812471343420449,
+      "grad_norm": 0.5540035367012024,
+      "learning_rate": 3.654730946421403e-05,
+      "loss": 2.0144,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8817056396148556,
+      "grad_norm": 0.5454185605049133,
+      "learning_rate": 3.6269091683204466e-05,
+      "loss": 1.9399,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8821641448876663,
+      "grad_norm": 0.5358849167823792,
+      "learning_rate": 3.5991897065521693e-05,
+      "loss": 2.0472,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8826226501604768,
+      "grad_norm": 0.5498891472816467,
+      "learning_rate": 3.571572622276026e-05,
+      "loss": 2.0238,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8830811554332875,
+      "grad_norm": 0.5495590567588806,
+      "learning_rate": 3.544057976425619e-05,
+      "loss": 2.0287,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8835396607060981,
+      "grad_norm": 0.5334488153457642,
+      "learning_rate": 3.5166458297085146e-05,
+      "loss": 2.0377,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8839981659789088,
+      "grad_norm": 0.5375003218650818,
+      "learning_rate": 3.489336242606111e-05,
+      "loss": 1.9752,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8844566712517193,
+      "grad_norm": 0.5299517512321472,
+      "learning_rate": 3.462129275373577e-05,
+      "loss": 1.9356,
+      "step": 1929
+    },
+    {
+      "epoch": 0.88491517652453,
+      "grad_norm": 0.5271756649017334,
+      "learning_rate": 3.4350249880395924e-05,
+      "loss": 2.0152,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8853736817973407,
+      "grad_norm": 0.5325633883476257,
+      "learning_rate": 3.408023440406355e-05,
+      "loss": 1.9428,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8858321870701513,
+      "grad_norm": 0.5418753623962402,
+      "learning_rate": 3.381124692049331e-05,
+      "loss": 2.0212,
+      "step": 1932
+    },
+    {
+      "epoch": 0.886290692342962,
+      "grad_norm": 0.5114116072654724,
+      "learning_rate": 3.354328802317197e-05,
+      "loss": 1.9405,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8867491976157725,
+      "grad_norm": 0.5555707216262817,
+      "learning_rate": 3.327635830331677e-05,
+      "loss": 1.9491,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8872077028885832,
+      "grad_norm": 0.5187960267066956,
+      "learning_rate": 3.3010458349874206e-05,
+      "loss": 1.9928,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8876662081613939,
+      "grad_norm": 0.5364302396774292,
+      "learning_rate": 3.2745588749518775e-05,
+      "loss": 1.9888,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8881247134342045,
+      "grad_norm": 0.5583979487419128,
+      "learning_rate": 3.248175008665161e-05,
+      "loss": 1.9096,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8885832187070152,
+      "grad_norm": 0.5368699431419373,
+      "learning_rate": 3.221894294339911e-05,
+      "loss": 1.9772,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8890417239798257,
+      "grad_norm": 0.5241401791572571,
+      "learning_rate": 3.1957167899611836e-05,
+      "loss": 1.9479,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8895002292526364,
+      "grad_norm": 0.5416818857192993,
+      "learning_rate": 3.169642553286334e-05,
+      "loss": 2.019,
+      "step": 1940
+    },
+    {
+      "epoch": 0.889958734525447,
+      "grad_norm": 0.5484623312950134,
+      "learning_rate": 3.143671641844831e-05,
+      "loss": 1.9978,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8904172397982577,
+      "grad_norm": 0.5441874265670776,
+      "learning_rate": 3.117804112938205e-05,
+      "loss": 1.932,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8908757450710684,
+      "grad_norm": 0.5218651294708252,
+      "learning_rate": 3.092040023639869e-05,
+      "loss": 1.9866,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8913342503438789,
+      "grad_norm": 0.5202623605728149,
+      "learning_rate": 3.066379430795002e-05,
+      "loss": 1.9364,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8917927556166896,
+      "grad_norm": 0.5251177549362183,
+      "learning_rate": 3.040822391020459e-05,
+      "loss": 2.0124,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8922512608895002,
+      "grad_norm": 0.5008144378662109,
+      "learning_rate": 3.0153689607045842e-05,
+      "loss": 1.9025,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8927097661623109,
+      "grad_norm": 0.5220935940742493,
+      "learning_rate": 2.9900191960071545e-05,
+      "loss": 1.9555,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8931682714351215,
+      "grad_norm": 0.536674976348877,
+      "learning_rate": 2.9647731528591848e-05,
+      "loss": 1.9578,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8936267767079321,
+      "grad_norm": 0.5261812210083008,
+      "learning_rate": 2.9396308869628795e-05,
+      "loss": 1.9463,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8940852819807428,
+      "grad_norm": 0.5679231286048889,
+      "learning_rate": 2.914592453791448e-05,
+      "loss": 1.9799,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8945437872535534,
+      "grad_norm": 0.5360503792762756,
+      "learning_rate": 2.8896579085889994e-05,
+      "loss": 1.9602,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8950022925263641,
+      "grad_norm": 0.5203515887260437,
+      "learning_rate": 2.86482730637046e-05,
+      "loss": 1.9369,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8954607977991746,
+      "grad_norm": 0.5228835344314575,
+      "learning_rate": 2.840100701921383e-05,
+      "loss": 1.97,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8959193030719853,
+      "grad_norm": 0.5458363890647888,
+      "learning_rate": 2.8154781497978898e-05,
+      "loss": 2.0231,
+      "step": 1954
+    },
+    {
+      "epoch": 0.896377808344796,
+      "grad_norm": 0.552959680557251,
+      "learning_rate": 2.7909597043265013e-05,
+      "loss": 1.9406,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8968363136176066,
+      "grad_norm": 0.5265349745750427,
+      "learning_rate": 2.7665454196040662e-05,
+      "loss": 1.9515,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8972948188904173,
+      "grad_norm": 0.5314714312553406,
+      "learning_rate": 2.7422353494975905e-05,
+      "loss": 1.9557,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8977533241632278,
+      "grad_norm": 0.5296239256858826,
+      "learning_rate": 2.7180295476441573e-05,
+      "loss": 1.9395,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8982118294360385,
+      "grad_norm": 0.5375972390174866,
+      "learning_rate": 2.6939280674508016e-05,
+      "loss": 1.9505,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8986703347088492,
+      "grad_norm": 0.5206801891326904,
+      "learning_rate": 2.669930962094358e-05,
+      "loss": 2.0012,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8991288399816598,
+      "grad_norm": 0.5384902954101562,
+      "learning_rate": 2.6460382845214126e-05,
+      "loss": 2.0651,
+      "step": 1961
+    },
+    {
+      "epoch": 0.8995873452544704,
+      "grad_norm": 0.5173669457435608,
+      "learning_rate": 2.6222500874481025e-05,
+      "loss": 2.0471,
+      "step": 1962
+    },
+    {
+      "epoch": 0.900045850527281,
+      "grad_norm": 0.5497618913650513,
+      "learning_rate": 2.5985664233600827e-05,
+      "loss": 1.9633,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9005043558000917,
+      "grad_norm": 0.5620598196983337,
+      "learning_rate": 2.574987344512336e-05,
+      "loss": 1.9479,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9009628610729024,
+      "grad_norm": 0.5087990164756775,
+      "learning_rate": 2.5515129029290984e-05,
+      "loss": 2.0159,
+      "step": 1965
+    },
+    {
+      "epoch": 0.901421366345713,
+      "grad_norm": 0.534902811050415,
+      "learning_rate": 2.5281431504037556e-05,
+      "loss": 1.975,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9018798716185236,
+      "grad_norm": 0.5336820483207703,
+      "learning_rate": 2.504878138498684e-05,
+      "loss": 1.9815,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9023383768913342,
+      "grad_norm": 0.542467474937439,
+      "learning_rate": 2.48171791854519e-05,
+      "loss": 1.9849,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9027968821641449,
+      "grad_norm": 0.5429170727729797,
+      "learning_rate": 2.4586625416433473e-05,
+      "loss": 1.9986,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9032553874369555,
+      "grad_norm": 0.526159405708313,
+      "learning_rate": 2.435712058661921e-05,
+      "loss": 2.0115,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9037138927097662,
+      "grad_norm": 0.5729237794876099,
+      "learning_rate": 2.4128665202382327e-05,
+      "loss": 1.9675,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9041723979825768,
+      "grad_norm": 0.5110265016555786,
+      "learning_rate": 2.3901259767780515e-05,
+      "loss": 1.9147,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9046309032553874,
+      "grad_norm": 0.5542974472045898,
+      "learning_rate": 2.367490478455514e-05,
+      "loss": 2.051,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9050894085281981,
+      "grad_norm": 0.5289274454116821,
+      "learning_rate": 2.3449600752129597e-05,
+      "loss": 1.9352,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9055479138010087,
+      "grad_norm": 0.5342292785644531,
+      "learning_rate": 2.3225348167608685e-05,
+      "loss": 1.9292,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9060064190738194,
+      "grad_norm": 0.5435636043548584,
+      "learning_rate": 2.3002147525777118e-05,
+      "loss": 2.0335,
+      "step": 1976
+    },
+    {
+      "epoch": 0.90646492434663,
+      "grad_norm": 0.5279362201690674,
+      "learning_rate": 2.2779999319098856e-05,
+      "loss": 1.9238,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9069234296194406,
+      "grad_norm": 0.5395612120628357,
+      "learning_rate": 2.255890403771571e-05,
+      "loss": 1.9845,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9073819348922513,
+      "grad_norm": 0.5196919441223145,
+      "learning_rate": 2.233886216944614e-05,
+      "loss": 1.9638,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9078404401650619,
+      "grad_norm": 0.5208037495613098,
+      "learning_rate": 2.211987419978484e-05,
+      "loss": 2.0352,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9082989454378725,
+      "grad_norm": 0.5482191443443298,
+      "learning_rate": 2.1901940611900705e-05,
+      "loss": 2.0601,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9087574507106831,
+      "grad_norm": 0.5299661159515381,
+      "learning_rate": 2.168506188663666e-05,
+      "loss": 1.9716,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9092159559834938,
+      "grad_norm": 0.5165415406227112,
+      "learning_rate": 2.1469238502507925e-05,
+      "loss": 1.9957,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9096744612563045,
+      "grad_norm": 0.5063280463218689,
+      "learning_rate": 2.125447093570154e-05,
+      "loss": 1.9511,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9101329665291151,
+      "grad_norm": 0.5495461225509644,
+      "learning_rate": 2.1040759660074793e-05,
+      "loss": 2.0016,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9105914718019257,
+      "grad_norm": 0.5401156544685364,
+      "learning_rate": 2.0828105147154273e-05,
+      "loss": 1.9487,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9110499770747363,
+      "grad_norm": 0.5542133450508118,
+      "learning_rate": 2.061650786613545e-05,
+      "loss": 1.9751,
+      "step": 1987
+    },
+    {
+      "epoch": 0.911508482347547,
+      "grad_norm": 0.5555367469787598,
+      "learning_rate": 2.040596828388058e-05,
+      "loss": 1.9837,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9119669876203577,
+      "grad_norm": 0.5437086224555969,
+      "learning_rate": 2.019648686491865e-05,
+      "loss": 1.9285,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9124254928931683,
+      "grad_norm": 0.5606649518013,
+      "learning_rate": 1.9988064071443767e-05,
+      "loss": 1.9054,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9128839981659789,
+      "grad_norm": 0.5466018915176392,
+      "learning_rate": 1.9780700363314253e-05,
+      "loss": 2.0224,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9133425034387895,
+      "grad_norm": 0.5408046841621399,
+      "learning_rate": 1.957439619805196e-05,
+      "loss": 1.9619,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9138010087116002,
+      "grad_norm": 0.5471007823944092,
+      "learning_rate": 1.9369152030840554e-05,
+      "loss": 1.9193,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9142595139844109,
+      "grad_norm": 0.540513813495636,
+      "learning_rate": 1.916496831452552e-05,
+      "loss": 1.9919,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9147180192572214,
+      "grad_norm": 0.5324971675872803,
+      "learning_rate": 1.8961845499611998e-05,
+      "loss": 1.8474,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9151765245300321,
+      "grad_norm": 0.5580838918685913,
+      "learning_rate": 1.8759784034264925e-05,
+      "loss": 2.0167,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9156350298028427,
+      "grad_norm": 0.5379783511161804,
+      "learning_rate": 1.855878436430708e-05,
+      "loss": 2.0264,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9160935350756534,
+      "grad_norm": 0.5449082851409912,
+      "learning_rate": 1.835884693321871e-05,
+      "loss": 1.9458,
+      "step": 1998
+    },
+    {
+      "epoch": 0.916552040348464,
+      "grad_norm": 0.5380805730819702,
+      "learning_rate": 1.8159972182136386e-05,
+      "loss": 2.0685,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9170105456212746,
+      "grad_norm": 0.5595501661300659,
+      "learning_rate": 1.7962160549851945e-05,
+      "loss": 1.9905,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9174690508940853,
+      "grad_norm": 0.5222651958465576,
+      "learning_rate": 1.776541247281177e-05,
+      "loss": 2.0018,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9179275561668959,
+      "grad_norm": 0.5367431640625,
+      "learning_rate": 1.7569728385115224e-05,
+      "loss": 1.9444,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9183860614397066,
+      "grad_norm": 0.5324277281761169,
+      "learning_rate": 1.7375108718514665e-05,
+      "loss": 1.9552,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9188445667125172,
+      "grad_norm": 0.5471982955932617,
+      "learning_rate": 1.7181553902413438e-05,
+      "loss": 1.9799,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9193030719853278,
+      "grad_norm": 0.530003547668457,
+      "learning_rate": 1.698906436386577e-05,
+      "loss": 1.987,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9197615772581385,
+      "grad_norm": 0.5258774161338806,
+      "learning_rate": 1.679764052757532e-05,
+      "loss": 2.0657,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9202200825309491,
+      "grad_norm": 0.5284263491630554,
+      "learning_rate": 1.6607282815894464e-05,
+      "loss": 1.917,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9206785878037598,
+      "grad_norm": 0.5248232483863831,
+      "learning_rate": 1.6417991648823405e-05,
+      "loss": 1.9378,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9211370930765704,
+      "grad_norm": 0.5525339841842651,
+      "learning_rate": 1.6229767444008835e-05,
+      "loss": 1.9467,
+      "step": 2009
+    },
+    {
+      "epoch": 0.921595598349381,
+      "grad_norm": 0.5393236875534058,
+      "learning_rate": 1.604261061674378e-05,
+      "loss": 1.9959,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9220541036221916,
+      "grad_norm": 0.5478919744491577,
+      "learning_rate": 1.5856521579965865e-05,
+      "loss": 1.9432,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9225126088950023,
+      "grad_norm": 0.5475273728370667,
+      "learning_rate": 1.5671500744256938e-05,
+      "loss": 2.0235,
+      "step": 2012
+    },
+    {
+      "epoch": 0.922971114167813,
+      "grad_norm": 0.561354398727417,
+      "learning_rate": 1.5487548517841953e-05,
+      "loss": 1.9182,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9234296194406235,
+      "grad_norm": 0.534937858581543,
+      "learning_rate": 1.530466530658814e-05,
+      "loss": 2.0139,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9238881247134342,
+      "grad_norm": 0.5717822909355164,
+      "learning_rate": 1.5122851514004054e-05,
+      "loss": 2.0365,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9243466299862448,
+      "grad_norm": 0.5567071437835693,
+      "learning_rate": 1.4942107541238703e-05,
+      "loss": 1.9984,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9248051352590555,
+      "grad_norm": 0.5186538100242615,
+      "learning_rate": 1.4762433787080809e-05,
+      "loss": 1.948,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9252636405318662,
+      "grad_norm": 0.5425750017166138,
+      "learning_rate": 1.4583830647957541e-05,
+      "loss": 2.0128,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9257221458046767,
+      "grad_norm": 0.5243472456932068,
+      "learning_rate": 1.4406298517934068e-05,
+      "loss": 1.8538,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9261806510774874,
+      "grad_norm": 0.5262764692306519,
+      "learning_rate": 1.4229837788712562e-05,
+      "loss": 1.8742,
+      "step": 2020
+    },
+    {
+      "epoch": 0.926639156350298,
+      "grad_norm": 0.5254957675933838,
+      "learning_rate": 1.4054448849631085e-05,
+      "loss": 2.0233,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9270976616231087,
+      "grad_norm": 0.5251672267913818,
+      "learning_rate": 1.3880132087663145e-05,
+      "loss": 1.937,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9275561668959194,
+      "grad_norm": 0.5195220708847046,
+      "learning_rate": 1.3706887887416419e-05,
+      "loss": 1.9625,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9280146721687299,
+      "grad_norm": 0.5315464735031128,
+      "learning_rate": 1.3534716631132316e-05,
+      "loss": 1.9175,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9284731774415406,
+      "grad_norm": 0.531907320022583,
+      "learning_rate": 1.3363618698684853e-05,
+      "loss": 1.9333,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9289316827143512,
+      "grad_norm": 0.5235713720321655,
+      "learning_rate": 1.3193594467579728e-05,
+      "loss": 1.9164,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9293901879871619,
+      "grad_norm": 0.5185490250587463,
+      "learning_rate": 1.3024644312954026e-05,
+      "loss": 1.9805,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9298486932599724,
+      "grad_norm": 0.5321975350379944,
+      "learning_rate": 1.2856768607574564e-05,
+      "loss": 1.9343,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9303071985327831,
+      "grad_norm": 0.5414285659790039,
+      "learning_rate": 1.2689967721837947e-05,
+      "loss": 2.0251,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9307657038055938,
+      "grad_norm": 0.5516597628593445,
+      "learning_rate": 1.2524242023769006e-05,
+      "loss": 1.9934,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9312242090784044,
+      "grad_norm": 0.5391863584518433,
+      "learning_rate": 1.2359591879020526e-05,
+      "loss": 1.973,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9316827143512151,
+      "grad_norm": 0.5544226169586182,
+      "learning_rate": 1.2196017650872081e-05,
+      "loss": 1.9646,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9321412196240256,
+      "grad_norm": 0.5757526159286499,
+      "learning_rate": 1.2033519700229367e-05,
+      "loss": 1.8858,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9325997248968363,
+      "grad_norm": 0.5463371872901917,
+      "learning_rate": 1.1872098385623586e-05,
+      "loss": 2.0406,
+      "step": 2034
+    },
+    {
+      "epoch": 0.933058230169647,
+      "grad_norm": 0.52485591173172,
+      "learning_rate": 1.1711754063210289e-05,
+      "loss": 1.9778,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9335167354424576,
+      "grad_norm": 0.5307138562202454,
+      "learning_rate": 1.155248708676887e-05,
+      "loss": 1.8767,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9339752407152683,
+      "grad_norm": 0.5389659404754639,
+      "learning_rate": 1.1394297807701737e-05,
+      "loss": 1.9466,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9344337459880788,
+      "grad_norm": 0.5353839993476868,
+      "learning_rate": 1.1237186575033254e-05,
+      "loss": 2.0217,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9348922512608895,
+      "grad_norm": 0.5359538197517395,
+      "learning_rate": 1.1081153735409522e-05,
+      "loss": 1.9335,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9353507565337001,
+      "grad_norm": 0.5385793447494507,
+      "learning_rate": 1.0926199633097156e-05,
+      "loss": 2.0128,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9358092618065108,
+      "grad_norm": 0.5385074019432068,
+      "learning_rate": 1.0772324609982787e-05,
+      "loss": 1.9885,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9362677670793215,
+      "grad_norm": 0.5435727834701538,
+      "learning_rate": 1.0619529005571893e-05,
+      "loss": 1.9753,
+      "step": 2042
+    },
+    {
+      "epoch": 0.936726272352132,
+      "grad_norm": 0.5461769700050354,
+      "learning_rate": 1.0467813156988748e-05,
+      "loss": 1.9835,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9371847776249427,
+      "grad_norm": 0.5254791378974915,
+      "learning_rate": 1.0317177398975031e-05,
+      "loss": 1.9523,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9376432828977533,
+      "grad_norm": 0.530639111995697,
+      "learning_rate": 1.0167622063889326e-05,
+      "loss": 1.9021,
+      "step": 2045
+    },
+    {
+      "epoch": 0.938101788170564,
+      "grad_norm": 0.5310068130493164,
+      "learning_rate": 1.0019147481706625e-05,
+      "loss": 1.9555,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9385602934433745,
+      "grad_norm": 0.547718346118927,
+      "learning_rate": 9.871753980017051e-06,
+      "loss": 1.9446,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9390187987161852,
+      "grad_norm": 0.5338418483734131,
+      "learning_rate": 9.725441884025855e-06,
+      "loss": 1.989,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9394773039889959,
+      "grad_norm": 0.5373159646987915,
+      "learning_rate": 9.580211516551862e-06,
+      "loss": 1.9547,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9399358092618065,
+      "grad_norm": 0.5363244414329529,
+      "learning_rate": 9.436063198027589e-06,
+      "loss": 1.9724,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9403943145346172,
+      "grad_norm": 0.5506080389022827,
+      "learning_rate": 9.292997246497959e-06,
+      "loss": 2.0381,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9408528198074277,
+      "grad_norm": 0.5553380846977234,
+      "learning_rate": 9.151013977619693e-06,
+      "loss": 1.8965,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9413113250802384,
+      "grad_norm": 0.5361767411231995,
+      "learning_rate": 9.010113704661038e-06,
+      "loss": 1.9191,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9417698303530491,
+      "grad_norm": 0.5397168397903442,
+      "learning_rate": 8.870296738500316e-06,
+      "loss": 1.9486,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9422283356258597,
+      "grad_norm": 0.5210735201835632,
+      "learning_rate": 8.731563387626095e-06,
+      "loss": 2.0534,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9426868408986704,
+      "grad_norm": 0.5368708968162537,
+      "learning_rate": 8.59391395813569e-06,
+      "loss": 1.9625,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9431453461714809,
+      "grad_norm": 0.5306990146636963,
+      "learning_rate": 8.457348753735328e-06,
+      "loss": 1.8999,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9436038514442916,
+      "grad_norm": 0.5263240933418274,
+      "learning_rate": 8.321868075738593e-06,
+      "loss": 2.0197,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9440623567171023,
+      "grad_norm": 0.5562419295310974,
+      "learning_rate": 8.187472223066371e-06,
+      "loss": 1.9789,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9445208619899129,
+      "grad_norm": 0.5402674078941345,
+      "learning_rate": 8.054161492246136e-06,
+      "loss": 1.9308,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9449793672627235,
+      "grad_norm": 0.5328630805015564,
+      "learning_rate": 7.921936177411049e-06,
+      "loss": 2.0275,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9454378725355341,
+      "grad_norm": 0.5638577342033386,
+      "learning_rate": 7.790796570299463e-06,
+      "loss": 1.9589,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9458963778083448,
+      "grad_norm": 0.5168532133102417,
+      "learning_rate": 7.660742960254207e-06,
+      "loss": 1.9701,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9463548830811555,
+      "grad_norm": 0.5436511039733887,
+      "learning_rate": 7.531775634222138e-06,
+      "loss": 1.912,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9468133883539661,
+      "grad_norm": 0.551314115524292,
+      "learning_rate": 7.403894876753192e-06,
+      "loss": 1.9178,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9472718936267767,
+      "grad_norm": 0.5396276712417603,
+      "learning_rate": 7.277100970000061e-06,
+      "loss": 1.9577,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9477303988995873,
+      "grad_norm": 0.5336462259292603,
+      "learning_rate": 7.151394193717408e-06,
+      "loss": 1.9351,
+      "step": 2067
+    },
+    {
+      "epoch": 0.948188904172398,
+      "grad_norm": 0.5853549242019653,
+      "learning_rate": 7.026774825261151e-06,
+      "loss": 2.0238,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9486474094452086,
+      "grad_norm": 0.523476243019104,
+      "learning_rate": 6.903243139588233e-06,
+      "loss": 1.9301,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9491059147180193,
+      "grad_norm": 0.5460129976272583,
+      "learning_rate": 6.780799409255522e-06,
+      "loss": 2.0194,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9495644199908299,
+      "grad_norm": 0.5273712277412415,
+      "learning_rate": 6.659443904419637e-06,
+      "loss": 2.0354,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9500229252636405,
+      "grad_norm": 0.5369364023208618,
+      "learning_rate": 6.539176892836008e-06,
+      "loss": 1.9035,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9504814305364512,
+      "grad_norm": 0.521734893321991,
+      "learning_rate": 6.4199986398585375e-06,
+      "loss": 2.0384,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9509399358092618,
+      "grad_norm": 0.5469264984130859,
+      "learning_rate": 6.3019094084388884e-06,
+      "loss": 2.0018,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9513984410820725,
+      "grad_norm": 0.5324346423149109,
+      "learning_rate": 6.18490945912592e-06,
+      "loss": 1.9034,
+      "step": 2075
+    },
+    {
+      "epoch": 0.951856946354883,
+      "grad_norm": 0.5182361006736755,
+      "learning_rate": 6.068999050065249e-06,
+      "loss": 1.9611,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9523154516276937,
+      "grad_norm": 0.5246878862380981,
+      "learning_rate": 5.9541784369983586e-06,
+      "loss": 1.9134,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9527739569005044,
+      "grad_norm": 0.567868709564209,
+      "learning_rate": 5.840447873262433e-06,
+      "loss": 1.923,
+      "step": 2078
+    },
+    {
+      "epoch": 0.953232462173315,
+      "grad_norm": 0.5284212827682495,
+      "learning_rate": 5.727807609789471e-06,
+      "loss": 1.9965,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9536909674461256,
+      "grad_norm": 0.5387609601020813,
+      "learning_rate": 5.616257895105892e-06,
+      "loss": 1.9357,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9541494727189362,
+      "grad_norm": 0.5239283442497253,
+      "learning_rate": 5.505798975331933e-06,
+      "loss": 2.0565,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9546079779917469,
+      "grad_norm": 0.5611881613731384,
+      "learning_rate": 5.396431094181198e-06,
+      "loss": 2.0127,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9550664832645576,
+      "grad_norm": 0.5357159376144409,
+      "learning_rate": 5.288154492960107e-06,
+      "loss": 1.9579,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9555249885373682,
+      "grad_norm": 0.5476114749908447,
+      "learning_rate": 5.1809694105671155e-06,
+      "loss": 1.8537,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9559834938101788,
+      "grad_norm": 0.5384935736656189,
+      "learning_rate": 5.074876083492441e-06,
+      "loss": 1.9367,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9564419990829894,
+      "grad_norm": 0.5304427146911621,
+      "learning_rate": 4.96987474581767e-06,
+      "loss": 2.0119,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9569005043558001,
+      "grad_norm": 0.5632122159004211,
+      "learning_rate": 4.865965629214819e-06,
+      "loss": 1.8863,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9573590096286108,
+      "grad_norm": 0.5295789837837219,
+      "learning_rate": 4.763148962946218e-06,
+      "loss": 1.9721,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9578175149014214,
+      "grad_norm": 0.5277626514434814,
+      "learning_rate": 4.661424973863681e-06,
+      "loss": 2.0007,
+      "step": 2089
+    },
+    {
+      "epoch": 0.958276020174232,
+      "grad_norm": 0.5383623242378235,
+      "learning_rate": 4.560793886408398e-06,
+      "loss": 1.9127,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9587345254470426,
+      "grad_norm": 0.5552330613136292,
+      "learning_rate": 4.461255922609986e-06,
+      "loss": 2.037,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9591930307198533,
+      "grad_norm": 0.5634151697158813,
+      "learning_rate": 4.362811302086267e-06,
+      "loss": 1.9231,
+      "step": 2092
+    },
+    {
+      "epoch": 0.959651535992664,
+      "grad_norm": 0.5389592051506042,
+      "learning_rate": 4.265460242042885e-06,
+      "loss": 1.9883,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9601100412654745,
+      "grad_norm": 0.5289784073829651,
+      "learning_rate": 4.169202957272522e-06,
+      "loss": 1.9438,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9605685465382852,
+      "grad_norm": 0.5449013710021973,
+      "learning_rate": 4.074039660154738e-06,
+      "loss": 1.9185,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9610270518110958,
+      "grad_norm": 0.5399166941642761,
+      "learning_rate": 3.9799705606551325e-06,
+      "loss": 1.977,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9614855570839065,
+      "grad_norm": 0.5689626932144165,
+      "learning_rate": 3.886995866325294e-06,
+      "loss": 1.9813,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9619440623567171,
+      "grad_norm": 0.5595788955688477,
+      "learning_rate": 3.795115782302072e-06,
+      "loss": 1.9964,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9624025676295277,
+      "grad_norm": 0.5320740938186646,
+      "learning_rate": 3.704330511307197e-06,
+      "loss": 1.962,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9628610729023384,
+      "grad_norm": 0.5339487195014954,
+      "learning_rate": 3.614640253646828e-06,
+      "loss": 1.9619,
+      "step": 2100
+    },
+    {
+      "epoch": 0.963319578175149,
+      "grad_norm": 0.5414394736289978,
+      "learning_rate": 3.5260452072110594e-06,
+      "loss": 2.0315,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9637780834479597,
+      "grad_norm": 0.575330913066864,
+      "learning_rate": 3.4385455674737498e-06,
+      "loss": 1.9454,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9642365887207703,
+      "grad_norm": 0.5406596064567566,
+      "learning_rate": 3.3521415274915256e-06,
+      "loss": 1.9092,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9646950939935809,
+      "grad_norm": 0.5587471723556519,
+      "learning_rate": 3.2668332779041133e-06,
+      "loss": 1.9917,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9651535992663915,
+      "grad_norm": 0.5710352659225464,
+      "learning_rate": 3.1826210069332838e-06,
+      "loss": 1.9829,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9656121045392022,
+      "grad_norm": 0.5492319464683533,
+      "learning_rate": 3.0995049003826324e-06,
+      "loss": 2.0167,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9660706098120129,
+      "grad_norm": 0.5250155329704285,
+      "learning_rate": 3.017485141637355e-06,
+      "loss": 1.9206,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9665291150848234,
+      "grad_norm": 0.5360960960388184,
+      "learning_rate": 2.9365619116636376e-06,
+      "loss": 1.9348,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9669876203576341,
+      "grad_norm": 0.5197913646697998,
+      "learning_rate": 2.856735389008269e-06,
+      "loss": 1.9545,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9674461256304447,
+      "grad_norm": 0.530626654624939,
+      "learning_rate": 2.778005749798307e-06,
+      "loss": 1.9411,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9679046309032554,
+      "grad_norm": 0.5235245823860168,
+      "learning_rate": 2.700373167740744e-06,
+      "loss": 2.0313,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9683631361760661,
+      "grad_norm": 0.5674985647201538,
+      "learning_rate": 2.62383781412201e-06,
+      "loss": 1.9486,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9688216414488766,
+      "grad_norm": 0.5370426177978516,
+      "learning_rate": 2.5483998578076373e-06,
+      "loss": 1.9681,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9692801467216873,
+      "grad_norm": 0.5314510464668274,
+      "learning_rate": 2.4740594652418736e-06,
+      "loss": 1.9714,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9697386519944979,
+      "grad_norm": 0.5319973826408386,
+      "learning_rate": 2.4008168004472917e-06,
+      "loss": 1.9809,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9701971572673086,
+      "grad_norm": 0.5313822627067566,
+      "learning_rate": 2.3286720250246253e-06,
+      "loss": 1.9408,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9706556625401193,
+      "grad_norm": 0.5480673313140869,
+      "learning_rate": 2.2576252981520994e-06,
+      "loss": 1.9663,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9711141678129298,
+      "grad_norm": 0.5133749842643738,
+      "learning_rate": 2.1876767765853233e-06,
+      "loss": 1.8897,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9715726730857405,
+      "grad_norm": 0.5351123809814453,
+      "learning_rate": 2.118826614656788e-06,
+      "loss": 1.9381,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9720311783585511,
+      "grad_norm": 0.5605373382568359,
+      "learning_rate": 2.051074964275701e-06,
+      "loss": 1.9409,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9724896836313618,
+      "grad_norm": 0.5516522526741028,
+      "learning_rate": 1.984421974927375e-06,
+      "loss": 1.9539,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9729481889041725,
+      "grad_norm": 0.5515019297599792,
+      "learning_rate": 1.9188677936731734e-06,
+      "loss": 2.0305,
+      "step": 2122
+    },
+    {
+      "epoch": 0.973406694176983,
+      "grad_norm": 0.5504806637763977,
+      "learning_rate": 1.8544125651501208e-06,
+      "loss": 2.0013,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9738651994497937,
+      "grad_norm": 0.5453490018844604,
+      "learning_rate": 1.7910564315704035e-06,
+      "loss": 1.9739,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9743237047226043,
+      "grad_norm": 0.5374801158905029,
+      "learning_rate": 1.7287995327214257e-06,
+      "loss": 1.9707,
+      "step": 2125
+    },
+    {
+      "epoch": 0.974782209995415,
+      "grad_norm": 0.5400498509407043,
+      "learning_rate": 1.6676420059649754e-06,
+      "loss": 1.9642,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9752407152682255,
+      "grad_norm": 0.5183267593383789,
+      "learning_rate": 1.6075839862374485e-06,
+      "loss": 1.8959,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9756992205410362,
+      "grad_norm": 0.5334120988845825,
+      "learning_rate": 1.5486256060492366e-06,
+      "loss": 1.9319,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9761577258138469,
+      "grad_norm": 0.539889931678772,
+      "learning_rate": 1.4907669954844495e-06,
+      "loss": 1.8677,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9766162310866575,
+      "grad_norm": 0.5138183236122131,
+      "learning_rate": 1.434008282200805e-06,
+      "loss": 1.947,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9770747363594682,
+      "grad_norm": 0.5680721402168274,
+      "learning_rate": 1.3783495914291844e-06,
+      "loss": 2.019,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9775332416322787,
+      "grad_norm": 0.5334733128547668,
+      "learning_rate": 1.3237910459734104e-06,
+      "loss": 2.0423,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9779917469050894,
+      "grad_norm": 0.5179668664932251,
+      "learning_rate": 1.270332766210025e-06,
+      "loss": 2.0269,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9784502521779,
+      "grad_norm": 0.5467820763587952,
+      "learning_rate": 1.2179748700879012e-06,
+      "loss": 2.0073,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9789087574507107,
+      "grad_norm": 0.5458849668502808,
+      "learning_rate": 1.1667174731280205e-06,
+      "loss": 1.9999,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9793672627235214,
+      "grad_norm": 0.5379295349121094,
+      "learning_rate": 1.1165606884234182e-06,
+      "loss": 1.9528,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9798257679963319,
+      "grad_norm": 0.5547844171524048,
+      "learning_rate": 1.0675046266386268e-06,
+      "loss": 1.9679,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9802842732691426,
+      "grad_norm": 0.5692421197891235,
+      "learning_rate": 1.019549396009567e-06,
+      "loss": 2.0721,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9807427785419532,
+      "grad_norm": 0.5399606823921204,
+      "learning_rate": 9.726951023434348e-07,
+      "loss": 1.975,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9812012838147639,
+      "grad_norm": 0.5400664210319519,
+      "learning_rate": 9.269418490182591e-07,
+      "loss": 2.0442,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9816597890875745,
+      "grad_norm": 0.5351099967956543,
+      "learning_rate": 8.822897369827332e-07,
+      "loss": 1.976,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9821182943603851,
+      "grad_norm": 0.5233093500137329,
+      "learning_rate": 8.387388647561611e-07,
+      "loss": 1.9294,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9825767996331958,
+      "grad_norm": 0.5238915681838989,
+      "learning_rate": 7.962893284279016e-07,
+      "loss": 1.9063,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9830353049060064,
+      "grad_norm": 0.5727565288543701,
+      "learning_rate": 7.549412216574791e-07,
+      "loss": 2.0597,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9834938101788171,
+      "grad_norm": 0.5466613173484802,
+      "learning_rate": 7.146946356743067e-07,
+      "loss": 1.8999,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9839523154516276,
+      "grad_norm": 0.534436821937561,
+      "learning_rate": 6.755496592773524e-07,
+      "loss": 1.9266,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9844108207244383,
+      "grad_norm": 0.5462943911552429,
+      "learning_rate": 6.375063788349733e-07,
+      "loss": 2.0439,
+      "step": 2147
+    },
+    {
+      "epoch": 0.984869325997249,
+      "grad_norm": 0.5436690449714661,
+      "learning_rate": 6.005648782848594e-07,
+      "loss": 1.9191,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9853278312700596,
+      "grad_norm": 0.5571165680885315,
+      "learning_rate": 5.647252391337565e-07,
+      "loss": 2.0205,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9857863365428703,
+      "grad_norm": 0.5436297655105591,
+      "learning_rate": 5.299875404572441e-07,
+      "loss": 1.9488,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9862448418156808,
+      "grad_norm": 0.5152961015701294,
+      "learning_rate": 4.963518588996796e-07,
+      "loss": 1.9034,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9867033470884915,
+      "grad_norm": 0.5236839056015015,
+      "learning_rate": 4.638182686738657e-07,
+      "loss": 1.9079,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9871618523613022,
+      "grad_norm": 0.5211430191993713,
+      "learning_rate": 4.3238684156110543e-07,
+      "loss": 1.8799,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9876203576341128,
+      "grad_norm": 0.5654520392417908,
+      "learning_rate": 4.020576469108139e-07,
+      "loss": 1.9203,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9880788629069235,
+      "grad_norm": 0.5463891625404358,
+      "learning_rate": 3.7283075164046274e-07,
+      "loss": 1.9824,
+      "step": 2155
+    },
+    {
+      "epoch": 0.988537368179734,
+      "grad_norm": 0.5271873474121094,
+      "learning_rate": 3.4470622023557995e-07,
+      "loss": 1.9246,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9889958734525447,
+      "grad_norm": 0.5315970182418823,
+      "learning_rate": 3.176841147492504e-07,
+      "loss": 1.9271,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9894543787253554,
+      "grad_norm": 0.5366193056106567,
+      "learning_rate": 2.9176449480244895e-07,
+      "loss": 1.8858,
+      "step": 2158
+    },
+    {
+      "epoch": 0.989912883998166,
+      "grad_norm": 0.5614402294158936,
+      "learning_rate": 2.6694741758342967e-07,
+      "loss": 2.0088,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9903713892709766,
+      "grad_norm": 0.5276125073432922,
+      "learning_rate": 2.432329378478926e-07,
+      "loss": 1.9806,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9908298945437872,
+      "grad_norm": 0.5197196006774902,
+      "learning_rate": 2.2062110791892798e-07,
+      "loss": 1.9488,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9912883998165979,
+      "grad_norm": 0.5789982676506042,
+      "learning_rate": 1.9911197768662792e-07,
+      "loss": 1.9719,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9917469050894085,
+      "grad_norm": 0.5261623859405518,
+      "learning_rate": 1.7870559460814173e-07,
+      "loss": 2.0126,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9922054103622192,
+      "grad_norm": 0.5512431263923645,
+      "learning_rate": 1.5940200370750947e-07,
+      "loss": 1.9902,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9926639156350298,
+      "grad_norm": 0.5523988008499146,
+      "learning_rate": 1.4120124757577291e-07,
+      "loss": 2.0371,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9931224209078404,
+      "grad_norm": 0.5569248199462891,
+      "learning_rate": 1.2410336637047603e-07,
+      "loss": 1.8924,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9935809261806511,
+      "grad_norm": 0.5510152578353882,
+      "learning_rate": 1.081083978159425e-07,
+      "loss": 2.052,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9940394314534617,
+      "grad_norm": 0.5172340869903564,
+      "learning_rate": 9.321637720310915e-08,
+      "loss": 1.9183,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9944979367262724,
+      "grad_norm": 0.5722464919090271,
+      "learning_rate": 7.942733738924845e-08,
+      "loss": 1.9959,
+      "step": 2169
+    },
+    {
+      "epoch": 0.994956441999083,
+      "grad_norm": 0.5176474452018738,
+      "learning_rate": 6.6741308798135e-08,
+      "loss": 1.9414,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9954149472718936,
+      "grad_norm": 0.5514294505119324,
+      "learning_rate": 5.5158319419934546e-08,
+      "loss": 2.047,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9958734525447043,
+      "grad_norm": 0.5316492915153503,
+      "learning_rate": 4.4678394810981906e-08,
+      "loss": 1.9525,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9963319578175149,
+      "grad_norm": 0.5237235426902771,
+      "learning_rate": 3.5301558093947527e-08,
+      "loss": 2.0249,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9967904630903255,
+      "grad_norm": 0.5459381341934204,
+      "learning_rate": 2.7027829957559925e-08,
+      "loss": 1.989,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9972489683631361,
+      "grad_norm": 0.5324474573135376,
+      "learning_rate": 1.985722865682771e-08,
+      "loss": 1.9809,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9977074736359468,
+      "grad_norm": 0.5287396907806396,
+      "learning_rate": 1.3789770012762048e-08,
+      "loss": 1.9253,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9981659789087575,
+      "grad_norm": 0.5416738390922546,
+      "learning_rate": 8.825467412376665e-09,
+      "loss": 1.9749,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9986244841815681,
+      "grad_norm": 0.5533262491226196,
+      "learning_rate": 4.96433180879885e-09,
+      "loss": 1.9868,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9990829894543787,
+      "grad_norm": 0.554210901260376,
+      "learning_rate": 2.206371721158451e-09,
+      "loss": 1.9378,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9995414947271893,
+      "grad_norm": 0.533422589302063,
+      "learning_rate": 5.515932345323549e-10,
+      "loss": 1.9269,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5145041942596436,
+      "learning_rate": 0.0,
+      "loss": 1.9961,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0,
+      "step": 2181,
+      "total_flos": 8.045217014664397e+17,
+      "train_loss": 2.1934414989955924,
+      "train_runtime": 10522.0511,
+      "train_samples_per_second": 53.044,
+      "train_steps_per_second": 0.207
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2181,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1.0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.045217014664397e+17,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}