diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9107 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0144827586206897,
+  "eval_steps": 145,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.4522,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 1.6368393898010254,
+      "eval_runtime": 52.4631,
+      "eval_samples_per_second": 14.715,
+      "eval_steps_per_second": 7.358,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.7542,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5e-06,
+      "loss": 1.4392,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.2693,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5e-06,
+      "loss": 2.0255,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-06,
+      "loss": 1.8536,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5e-06,
+      "loss": 1.364,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.7387,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5e-06,
+      "loss": 1.6494,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-06,
+      "loss": 1.7298,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999998522885947e-06,
+      "loss": 1.1661,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999994091545531e-06,
+      "loss": 1.8255,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999986705983991e-06,
+      "loss": 0.8458,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999976366210054e-06,
+      "loss": 1.1704,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999963072235937e-06,
+      "loss": 1.8649,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.99994682407735e-06,
+      "loss": 1.2619,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999927621753493e-06,
+      "loss": 1.9507,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999905465287059e-06,
+      "loss": 1.4712,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999880354704227e-06,
+      "loss": 1.4141,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999852290034674e-06,
+      "loss": 0.9806,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9998212713115605e-06,
+      "loss": 1.0796,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999787298571541e-06,
+      "loss": 1.8835,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999750371854763e-06,
+      "loss": 1.3385,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9997104912048615e-06,
+      "loss": 1.2733,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999667656668961e-06,
+      "loss": 1.1878,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999621868297683e-06,
+      "loss": 1.3143,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999573126145132e-06,
+      "loss": 0.9616,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999521430268907e-06,
+      "loss": 1.5245,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999466780730097e-06,
+      "loss": 1.135,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999409177593279e-06,
+      "loss": 1.2351,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999348620926526e-06,
+      "loss": 1.3666,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999285110801393e-06,
+      "loss": 1.6391,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999218647292932e-06,
+      "loss": 1.564,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999149230479682e-06,
+      "loss": 1.129,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999076860443671e-06,
+      "loss": 1.3177,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999001537270419e-06,
+      "loss": 1.2801,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998923261048935e-06,
+      "loss": 1.4633,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998842031871717e-06,
+      "loss": 1.4125,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998757849834752e-06,
+      "loss": 1.0231,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998670715037518e-06,
+      "loss": 1.4911,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9985806275829815e-06,
+      "loss": 1.4786,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998487587577598e-06,
+      "loss": 1.4127,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9983915951313125e-06,
+      "loss": 0.8299,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998292650357558e-06,
+      "loss": 1.4409,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998190753373256e-06,
+      "loss": 1.3596,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998085904298817e-06,
+      "loss": 1.0493,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.997978103258142e-06,
+      "loss": 1.5774,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.997867350378618e-06,
+      "loss": 0.9767,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.99775364579112e-06,
+      "loss": 1.2138,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.997636989630012e-06,
+      "loss": 1.416,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997517382033145e-06,
+      "loss": 1.017,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9973948231418596e-06,
+      "loss": 1.6522,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997269313100981e-06,
+      "loss": 1.0402,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997140852058825e-06,
+      "loss": 1.1969,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997009440167192e-06,
+      "loss": 1.863,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.99687507758137e-06,
+      "loss": 1.3628,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996737764460135e-06,
+      "loss": 1.3267,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996597500965747e-06,
+      "loss": 1.8708,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996454287263955e-06,
+      "loss": 1.1691,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996308123523996e-06,
+      "loss": 1.5923,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996159009918586e-06,
+      "loss": 1.8901,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996006946623933e-06,
+      "loss": 1.6246,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.99585193381973e-06,
+      "loss": 1.5248,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.995693971689153e-06,
+      "loss": 1.405,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.995533060418865e-06,
+      "loss": 1.6614,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.995369200199014e-06,
+      "loss": 1.5643,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.995202391223231e-06,
+      "loss": 1.2362,
+      "step": 67
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.995032633688633e-06,
+      "loss": 1.0164,
+      "step": 68
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9948599277958224e-06,
+      "loss": 1.3667,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.994684273748881e-06,
+      "loss": 1.141,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9945056717553816e-06,
+      "loss": 0.8697,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.994324122026374e-06,
+      "loss": 1.1478,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.994139624776394e-06,
+      "loss": 0.9953,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.993952180223461e-06,
+      "loss": 0.9585,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.993761788589078e-06,
+      "loss": 1.583,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.993568450098227e-06,
+      "loss": 1.7872,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.993372164979375e-06,
+      "loss": 1.6221,
+      "step": 77
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99317293346447e-06,
+      "loss": 1.5637,
+      "step": 78
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.992970755788944e-06,
+      "loss": 0.9444,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.992765632191706e-06,
+      "loss": 1.3295,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.992557562915152e-06,
+      "loss": 1.2464,
+      "step": 81
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.992346548205152e-06,
+      "loss": 1.0352,
+      "step": 82
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.992132588311062e-06,
+      "loss": 1.0081,
+      "step": 83
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9919156834857165e-06,
+      "loss": 1.4097,
+      "step": 84
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.991695833985429e-06,
+      "loss": 1.3536,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.991473040069996e-06,
+      "loss": 1.4188,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.99124730200269e-06,
+      "loss": 0.9988,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.991018620050263e-06,
+      "loss": 1.4854,
+      "step": 88
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.990786994482947e-06,
+      "loss": 1.3899,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.990552425574452e-06,
+      "loss": 1.1053,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.990314913601966e-06,
+      "loss": 1.0796,
+      "step": 91
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.990074458846155e-06,
+      "loss": 1.3599,
+      "step": 92
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.989831061591161e-06,
+      "loss": 1.2571,
+      "step": 93
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.989584722124606e-06,
+      "loss": 1.6078,
+      "step": 94
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.989335440737587e-06,
+      "loss": 1.09,
+      "step": 95
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.989083217724676e-06,
+      "loss": 1.3914,
+      "step": 96
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.988828053383925e-06,
+      "loss": 1.0266,
+      "step": 97
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.988569948016858e-06,
+      "loss": 1.5676,
+      "step": 98
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.988308901928477e-06,
+      "loss": 1.0114,
+      "step": 99
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.988044915427257e-06,
+      "loss": 1.0441,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.987777988825147e-06,
+      "loss": 1.1977,
+      "step": 101
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.987508122437575e-06,
+      "loss": 1.0235,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.987235316583439e-06,
+      "loss": 1.886,
+      "step": 103
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.986959571585109e-06,
+      "loss": 0.8832,
+      "step": 104
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.986680887768433e-06,
+      "loss": 2.0845,
+      "step": 105
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.986399265462728e-06,
+      "loss": 1.2562,
+      "step": 106
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.986114705000785e-06,
+      "loss": 0.951,
+      "step": 107
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.985827206718866e-06,
+      "loss": 0.9751,
+      "step": 108
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.985536770956706e-06,
+      "loss": 1.048,
+      "step": 109
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9852433980575095e-06,
+      "loss": 1.5994,
+      "step": 110
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.984947088367954e-06,
+      "loss": 1.2163,
+      "step": 111
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.984647842238185e-06,
+      "loss": 1.2535,
+      "step": 112
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.984345660021819e-06,
+      "loss": 1.0773,
+      "step": 113
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.984040542075942e-06,
+      "loss": 1.7597,
+      "step": 114
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.98373248876111e-06,
+      "loss": 0.9974,
+      "step": 115
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9834215004413474e-06,
+      "loss": 1.6053,
+      "step": 116
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.983107577484145e-06,
+      "loss": 1.0382,
+      "step": 117
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.982790720260463e-06,
+      "loss": 1.4656,
+      "step": 118
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.982470929144729e-06,
+      "loss": 1.187,
+      "step": 119
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.982148204514838e-06,
+      "loss": 0.9327,
+      "step": 120
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.98182254675215e-06,
+      "loss": 1.168,
+      "step": 121
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.981493956241492e-06,
+      "loss": 1.1143,
+      "step": 122
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.981162433371157e-06,
+      "loss": 1.6322,
+      "step": 123
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.980827978532903e-06,
+      "loss": 1.4653,
+      "step": 124
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.980490592121951e-06,
+      "loss": 1.4012,
+      "step": 125
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.980150274536989e-06,
+      "loss": 1.7911,
+      "step": 126
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9798070261801665e-06,
+      "loss": 1.5263,
+      "step": 127
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.979460847457097e-06,
+      "loss": 1.3853,
+      "step": 128
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9791117387768575e-06,
+      "loss": 1.3603,
+      "step": 129
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.978759700551986e-06,
+      "loss": 1.1095,
+      "step": 130
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.978404733198483e-06,
+      "loss": 0.8909,
+      "step": 131
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9780468371358116e-06,
+      "loss": 1.0373,
+      "step": 132
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.977686012786893e-06,
+      "loss": 0.9252,
+      "step": 133
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.97732226057811e-06,
+      "loss": 1.7641,
+      "step": 134
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.976955580939306e-06,
+      "loss": 1.4047,
+      "step": 135
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.976585974303785e-06,
+      "loss": 1.0995,
+      "step": 136
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.976213441108304e-06,
+      "loss": 1.1907,
+      "step": 137
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.975837981793085e-06,
+      "loss": 1.0744,
+      "step": 138
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.975459596801805e-06,
+      "loss": 1.097,
+      "step": 139
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.975078286581596e-06,
+      "loss": 0.8977,
+      "step": 140
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9746940515830514e-06,
+      "loss": 0.5722,
+      "step": 141
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.974306892260217e-06,
+      "loss": 1.455,
+      "step": 142
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.973916809070597e-06,
+      "loss": 1.6582,
+      "step": 143
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.973523802475147e-06,
+      "loss": 1.1232,
+      "step": 144
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9731278729382814e-06,
+      "loss": 1.1224,
+      "step": 145
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 1.3059237003326416,
+      "eval_runtime": 53.7489,
+      "eval_samples_per_second": 14.363,
+      "eval_steps_per_second": 7.182,
+      "step": 145
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.972729020927866e-06,
+      "loss": 1.1827,
+      "step": 146
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.97232724691522e-06,
+      "loss": 0.8917,
+      "step": 147
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.971922551375117e-06,
+      "loss": 0.8041,
+      "step": 148
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.971514934785783e-06,
+      "loss": 1.5842,
+      "step": 149
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9711043976288935e-06,
+      "loss": 1.5451,
+      "step": 150
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.970690940389576e-06,
+      "loss": 0.7718,
+      "step": 151
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.970274563556412e-06,
+      "loss": 1.3965,
+      "step": 152
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.969855267621428e-06,
+      "loss": 0.7756,
+      "step": 153
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.969433053080105e-06,
+      "loss": 1.5195,
+      "step": 154
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.969007920431367e-06,
+      "loss": 1.2027,
+      "step": 155
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.968579870177591e-06,
+      "loss": 1.7006,
+      "step": 156
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9681489028246e-06,
+      "loss": 1.8294,
+      "step": 157
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9677150188816655e-06,
+      "loss": 1.1739,
+      "step": 158
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.967278218861502e-06,
+      "loss": 1.4111,
+      "step": 159
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.966838503280275e-06,
+      "loss": 1.4639,
+      "step": 160
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.966395872657591e-06,
+      "loss": 1.4884,
+      "step": 161
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.965950327516502e-06,
+      "loss": 1.5146,
+      "step": 162
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.965501868383507e-06,
+      "loss": 1.2453,
+      "step": 163
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.965050495788544e-06,
+      "loss": 1.4534,
+      "step": 164
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.964596210264997e-06,
+      "loss": 1.0392,
+      "step": 165
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9641390123496926e-06,
+      "loss": 1.6082,
+      "step": 166
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9636789025828945e-06,
+      "loss": 1.2436,
+      "step": 167
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9632158815083135e-06,
+      "loss": 1.4229,
+      "step": 168
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9627499496730945e-06,
+      "loss": 1.5076,
+      "step": 169
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.962281107627828e-06,
+      "loss": 1.1288,
+      "step": 170
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.961809355926539e-06,
+      "loss": 0.9409,
+      "step": 171
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.961334695126693e-06,
+      "loss": 0.944,
+      "step": 172
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.960857125789191e-06,
+      "loss": 1.1763,
+      "step": 173
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9603766484783745e-06,
+      "loss": 1.0742,
+      "step": 174
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9598932637620175e-06,
+      "loss": 1.4498,
+      "step": 175
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.959406972211334e-06,
+      "loss": 1.8762,
+      "step": 176
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.958917774400967e-06,
+      "loss": 0.7468,
+      "step": 177
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.958425670909001e-06,
+      "loss": 1.8263,
+      "step": 178
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.957930662316948e-06,
+      "loss": 0.9994,
+      "step": 179
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.957432749209755e-06,
+      "loss": 1.1363,
+      "step": 180
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.956931932175803e-06,
+      "loss": 1.0489,
+      "step": 181
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.956428211806903e-06,
+      "loss": 0.9684,
+      "step": 182
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.955921588698295e-06,
+      "loss": 1.2817,
+      "step": 183
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.955412063448654e-06,
+      "loss": 1.5433,
+      "step": 184
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.95489963666008e-06,
+      "loss": 1.243,
+      "step": 185
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.954384308938102e-06,
+      "loss": 1.4317,
+      "step": 186
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.953866080891682e-06,
+      "loss": 0.8239,
+      "step": 187
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.953344953133201e-06,
+      "loss": 1.2553,
+      "step": 188
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.952820926278474e-06,
+      "loss": 1.1115,
+      "step": 189
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.952294000946738e-06,
+      "loss": 1.1013,
+      "step": 190
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.951764177760657e-06,
+      "loss": 0.9678,
+      "step": 191
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.951231457346317e-06,
+      "loss": 0.8286,
+      "step": 192
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.95069584033323e-06,
+      "loss": 1.0927,
+      "step": 193
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9501573273543295e-06,
+      "loss": 1.046,
+      "step": 194
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.949615919045972e-06,
+      "loss": 1.0288,
+      "step": 195
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.949071616047935e-06,
+      "loss": 1.5807,
+      "step": 196
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.948524419003415e-06,
+      "loss": 1.2632,
+      "step": 197
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.947974328559032e-06,
+      "loss": 1.5379,
+      "step": 198
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.947421345364822e-06,
+      "loss": 1.5668,
+      "step": 199
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.946865470074241e-06,
+      "loss": 1.1485,
+      "step": 200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.946306703344161e-06,
+      "loss": 1.1301,
+      "step": 201
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9457450458348735e-06,
+      "loss": 1.9502,
+      "step": 202
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.945180498210083e-06,
+      "loss": 1.056,
+      "step": 203
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.94461306113691e-06,
+      "loss": 1.1029,
+      "step": 204
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.944042735285891e-06,
+      "loss": 0.9279,
+      "step": 205
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.943469521330975e-06,
+      "loss": 1.1618,
+      "step": 206
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.942893419949523e-06,
+      "loss": 1.1059,
+      "step": 207
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9423144318223085e-06,
+      "loss": 2.1417,
+      "step": 208
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.941732557633519e-06,
+      "loss": 1.2264,
+      "step": 209
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.941147798070748e-06,
+      "loss": 1.1846,
+      "step": 210
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.940560153825003e-06,
+      "loss": 1.1038,
+      "step": 211
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.939969625590696e-06,
+      "loss": 1.6479,
+      "step": 212
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9393762140656484e-06,
+      "loss": 1.4235,
+      "step": 213
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.938779919951092e-06,
+      "loss": 1.0992,
+      "step": 214
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.938180743951661e-06,
+      "loss": 1.2565,
+      "step": 215
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.937578686775396e-06,
+      "loss": 1.3175,
+      "step": 216
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.936973749133743e-06,
+      "loss": 0.8759,
+      "step": 217
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.936365931741553e-06,
+      "loss": 0.8367,
+      "step": 218
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9357552353170755e-06,
+      "loss": 1.0833,
+      "step": 219
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.935141660581967e-06,
+      "loss": 1.2446,
+      "step": 220
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.934525208261284e-06,
+      "loss": 1.5117,
+      "step": 221
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.933905879083481e-06,
+      "loss": 1.0355,
+      "step": 222
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.933283673780416e-06,
+      "loss": 0.8376,
+      "step": 223
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.932658593087341e-06,
+      "loss": 1.0228,
+      "step": 224
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.93203063774291e-06,
+      "loss": 1.3021,
+      "step": 225
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9313998084891734e-06,
+      "loss": 1.3589,
+      "step": 226
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.930766106071574e-06,
+      "loss": 1.1009,
+      "step": 227
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.930129531238955e-06,
+      "loss": 1.0263,
+      "step": 228
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.929490084743549e-06,
+      "loss": 1.0732,
+      "step": 229
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.928847767340985e-06,
+      "loss": 1.6764,
+      "step": 230
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.928202579790285e-06,
+      "loss": 1.3953,
+      "step": 231
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.92755452285386e-06,
+      "loss": 1.3431,
+      "step": 232
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.926903597297514e-06,
+      "loss": 1.4097,
+      "step": 233
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.926249803890441e-06,
+      "loss": 1.0773,
+      "step": 234
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.925593143405221e-06,
+      "loss": 1.1947,
+      "step": 235
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.924933616617824e-06,
+      "loss": 1.6708,
+      "step": 236
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.924271224307609e-06,
+      "loss": 1.5323,
+      "step": 237
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9236059672573175e-06,
+      "loss": 1.4578,
+      "step": 238
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.922937846253078e-06,
+      "loss": 0.4763,
+      "step": 239
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.922266862084405e-06,
+      "loss": 0.9389,
+      "step": 240
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.921593015544193e-06,
+      "loss": 1.1362,
+      "step": 241
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9209163074287195e-06,
+      "loss": 1.3309,
+      "step": 242
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.920236738537647e-06,
+      "loss": 1.3693,
+      "step": 243
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.919554309674015e-06,
+      "loss": 1.0019,
+      "step": 244
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.918869021644243e-06,
+      "loss": 0.9216,
+      "step": 245
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9181808752581316e-06,
+      "loss": 1.089,
+      "step": 246
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.917489871328856e-06,
+      "loss": 2.1136,
+      "step": 247
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.916796010672969e-06,
+      "loss": 1.3636,
+      "step": 248
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.916099294110401e-06,
+      "loss": 1.6074,
+      "step": 249
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.915399722464455e-06,
+      "loss": 1.046,
+      "step": 250
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.91469729656181e-06,
+      "loss": 1.705,
+      "step": 251
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.913992017232515e-06,
+      "loss": 0.9568,
+      "step": 252
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.913283885309994e-06,
+      "loss": 1.0197,
+      "step": 253
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.912572901631038e-06,
+      "loss": 1.2031,
+      "step": 254
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.911859067035812e-06,
+      "loss": 1.5773,
+      "step": 255
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.911142382367847e-06,
+      "loss": 0.8763,
+      "step": 256
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.910422848474044e-06,
+      "loss": 2.1782,
+      "step": 257
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.90970046620467e-06,
+      "loss": 1.8428,
+      "step": 258
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.908975236413356e-06,
+      "loss": 0.7872,
+      "step": 259
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.908247159957102e-06,
+      "loss": 1.5376,
+      "step": 260
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.907516237696268e-06,
+      "loss": 1.1867,
+      "step": 261
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.906782470494579e-06,
+      "loss": 0.8102,
+      "step": 262
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.906045859219122e-06,
+      "loss": 0.8138,
+      "step": 263
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.905306404740342e-06,
+      "loss": 1.2403,
+      "step": 264
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.904564107932048e-06,
+      "loss": 1.4051,
+      "step": 265
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.903818969671404e-06,
+      "loss": 1.4872,
+      "step": 266
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.903070990838936e-06,
+      "loss": 1.8003,
+      "step": 267
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.902320172318521e-06,
+      "loss": 1.0195,
+      "step": 268
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.901566514997396e-06,
+      "loss": 0.814,
+      "step": 269
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.900810019766151e-06,
+      "loss": 0.9092,
+      "step": 270
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.90005068751873e-06,
+      "loss": 1.1939,
+      "step": 271
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.89928851915243e-06,
+      "loss": 0.939,
+      "step": 272
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.898523515567897e-06,
+      "loss": 1.079,
+      "step": 273
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.897755677669131e-06,
+      "loss": 0.5944,
+      "step": 274
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.896985006363477e-06,
+      "loss": 1.3437,
+      "step": 275
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.896211502561633e-06,
+      "loss": 1.5517,
+      "step": 276
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.89543516717764e-06,
+      "loss": 1.1935,
+      "step": 277
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.894656001128887e-06,
+      "loss": 1.1588,
+      "step": 278
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.893874005336108e-06,
+      "loss": 1.5299,
+      "step": 279
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.89308918072338e-06,
+      "loss": 1.9043,
+      "step": 280
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.892301528218124e-06,
+      "loss": 1.2985,
+      "step": 281
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.891511048751102e-06,
+      "loss": 1.4682,
+      "step": 282
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8907177432564165e-06,
+      "loss": 1.3038,
+      "step": 283
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.889921612671511e-06,
+      "loss": 1.5238,
+      "step": 284
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.889122657937163e-06,
+      "loss": 1.4669,
+      "step": 285
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.888320879997493e-06,
+      "loss": 1.3409,
+      "step": 286
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8875162797999545e-06,
+      "loss": 1.1065,
+      "step": 287
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.886708858295335e-06,
+      "loss": 1.2819,
+      "step": 288
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.885898616437759e-06,
+      "loss": 1.1449,
+      "step": 289
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.885085555184681e-06,
+      "loss": 1.3072,
+      "step": 290
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 1.2836387157440186,
+      "eval_runtime": 53.6026,
+      "eval_samples_per_second": 14.402,
+      "eval_steps_per_second": 7.201,
+      "step": 290
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.88426967549689e-06,
+      "loss": 0.6308,
+      "step": 291
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.883450978338503e-06,
+      "loss": 1.0291,
+      "step": 292
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.882629464676967e-06,
+      "loss": 1.1062,
+      "step": 293
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.881805135483057e-06,
+      "loss": 1.2533,
+      "step": 294
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.880977991730877e-06,
+      "loss": 1.2679,
+      "step": 295
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.880148034397854e-06,
+      "loss": 1.4962,
+      "step": 296
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.879315264464744e-06,
+      "loss": 1.0716,
+      "step": 297
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.87847968291562e-06,
+      "loss": 1.023,
+      "step": 298
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 1.3586,
+      "step": 299
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.876800088922257e-06,
+      "loss": 1.0656,
+      "step": 300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.875956078462778e-06,
+      "loss": 1.8788,
+      "step": 301
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.875109260356809e-06,
+      "loss": 1.0744,
+      "step": 302
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.874259635605025e-06,
+      "loss": 0.928,
+      "step": 303
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.8734072052114215e-06,
+      "loss": 0.9866,
+      "step": 304
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.872551970183309e-06,
+      "loss": 1.0274,
+      "step": 305
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.871693931531309e-06,
+      "loss": 1.0472,
+      "step": 306
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.87083309026936e-06,
+      "loss": 0.915,
+      "step": 307
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.86996944741471e-06,
+      "loss": 1.5422,
+      "step": 308
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.869103003987919e-06,
+      "loss": 1.2316,
+      "step": 309
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.868233761012853e-06,
+      "loss": 1.7933,
+      "step": 310
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.867361719516692e-06,
+      "loss": 0.7911,
+      "step": 311
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.866486880529918e-06,
+      "loss": 1.6406,
+      "step": 312
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.865609245086321e-06,
+      "loss": 1.2438,
+      "step": 313
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.864728814222996e-06,
+      "loss": 1.3507,
+      "step": 314
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.863845588980338e-06,
+      "loss": 1.355,
+      "step": 315
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.86295957040205e-06,
+      "loss": 0.9027,
+      "step": 316
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.862070759535129e-06,
+      "loss": 1.2907,
+      "step": 317
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.861179157429877e-06,
+      "loss": 1.9344,
+      "step": 318
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.860284765139892e-06,
+      "loss": 1.7715,
+      "step": 319
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.859387583722069e-06,
+      "loss": 1.1291,
+      "step": 320
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8584876142366e-06,
+      "loss": 0.9811,
+      "step": 321
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8575848577469716e-06,
+      "loss": 1.0526,
+      "step": 322
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.856679315319962e-06,
+      "loss": 0.8678,
+      "step": 323
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.855770988025644e-06,
+      "loss": 1.5962,
+      "step": 324
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.854859876937379e-06,
+      "loss": 1.9452,
+      "step": 325
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.853945983131819e-06,
+      "loss": 1.3036,
+      "step": 326
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.853029307688905e-06,
+      "loss": 1.0005,
+      "step": 327
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.852109851691864e-06,
+      "loss": 1.3575,
+      "step": 328
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.851187616227209e-06,
+      "loss": 0.9331,
+      "step": 329
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.850262602384738e-06,
+      "loss": 1.049,
+      "step": 330
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8493348112575315e-06,
+      "loss": 1.3,
+      "step": 331
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.848404243941951e-06,
+      "loss": 1.3482,
+      "step": 332
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.847470901537642e-06,
+      "loss": 0.8853,
+      "step": 333
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.846534785147526e-06,
+      "loss": 0.9279,
+      "step": 334
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.845595895877802e-06,
+      "loss": 1.2538,
+      "step": 335
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8446542348379495e-06,
+      "loss": 0.9986,
+      "step": 336
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.84370980314072e-06,
+      "loss": 1.4702,
+      "step": 337
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.842762601902141e-06,
+      "loss": 1.3831,
+      "step": 338
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.84181263224151e-06,
+      "loss": 1.1784,
+      "step": 339
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.840859895281401e-06,
+      "loss": 1.5826,
+      "step": 340
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.839904392147652e-06,
+      "loss": 0.8636,
+      "step": 341
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.838946123969373e-06,
+      "loss": 1.5024,
+      "step": 342
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8379850918789435e-06,
+      "loss": 0.9235,
+      "step": 343
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.837021297012004e-06,
+      "loss": 1.5189,
+      "step": 344
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.836054740507464e-06,
+      "loss": 1.3184,
+      "step": 345
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8350854235074945e-06,
+      "loss": 1.0936,
+      "step": 346
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.834113347157529e-06,
+      "loss": 1.2336,
+      "step": 347
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.83313851260626e-06,
+      "loss": 1.4692,
+      "step": 348
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.832160921005644e-06,
+      "loss": 1.3801,
+      "step": 349
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.83118057351089e-06,
+      "loss": 1.228,
+      "step": 350
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.830197471280467e-06,
+      "loss": 1.0152,
+      "step": 351
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.829211615476098e-06,
+      "loss": 0.9074,
+      "step": 352
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.82822300726276e-06,
+      "loss": 1.2977,
+      "step": 353
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.827231647808683e-06,
+      "loss": 1.1979,
+      "step": 354
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8262375382853485e-06,
+      "loss": 1.011,
+      "step": 355
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.825240679867486e-06,
+      "loss": 1.5818,
+      "step": 356
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8242410737330745e-06,
+      "loss": 1.1383,
+      "step": 357
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.82323872106334e-06,
+      "loss": 1.4268,
+      "step": 358
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.822233623042754e-06,
+      "loss": 0.779,
+      "step": 359
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.821225780859032e-06,
+      "loss": 1.3415,
+      "step": 360
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8202151957031316e-06,
+      "loss": 0.931,
+      "step": 361
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.819201868769253e-06,
+      "loss": 1.416,
+      "step": 362
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.818185801254837e-06,
+      "loss": 1.5162,
+      "step": 363
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.817166994360559e-06,
+      "loss": 1.4471,
+      "step": 364
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.816145449290336e-06,
+      "loss": 1.5774,
+      "step": 365
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.815121167251318e-06,
+      "loss": 1.3644,
+      "step": 366
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.814094149453891e-06,
+      "loss": 1.7256,
+      "step": 367
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.813064397111673e-06,
+      "loss": 1.6978,
+      "step": 368
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.812031911441512e-06,
+      "loss": 1.3279,
+      "step": 369
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8109966936634885e-06,
+      "loss": 1.4012,
+      "step": 370
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.80995874500091e-06,
+      "loss": 1.2934,
+      "step": 371
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.808918066680311e-06,
+      "loss": 1.1003,
+      "step": 372
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.807874659931452e-06,
+      "loss": 1.2407,
+      "step": 373
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.806828525987319e-06,
+      "loss": 0.7935,
+      "step": 374
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8057796660841175e-06,
+      "loss": 1.0551,
+      "step": 375
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.804728081461276e-06,
+      "loss": 1.3337,
+      "step": 376
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.803673773361444e-06,
+      "loss": 1.2896,
+      "step": 377
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.802616743030488e-06,
+      "loss": 1.2686,
+      "step": 378
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8015569917174905e-06,
+      "loss": 1.81,
+      "step": 379
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.800494520674752e-06,
+      "loss": 0.8058,
+      "step": 380
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.7994293311577835e-06,
+      "loss": 1.3053,
+      "step": 381
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.798361424425312e-06,
+      "loss": 1.3472,
+      "step": 382
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.7972908017392715e-06,
+      "loss": 1.6796,
+      "step": 383
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.796217464364808e-06,
+      "loss": 1.008,
+      "step": 384
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.795141413570275e-06,
+      "loss": 0.7311,
+      "step": 385
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.794062650627234e-06,
+      "loss": 1.1742,
+      "step": 386
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.792981176810447e-06,
+      "loss": 1.1034,
+      "step": 387
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.791896993397885e-06,
+      "loss": 0.8392,
+      "step": 388
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.790810101670715e-06,
+      "loss": 1.1105,
+      "step": 389
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.789720502913308e-06,
+      "loss": 1.3574,
+      "step": 390
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.788628198413236e-06,
+      "loss": 1.4231,
+      "step": 391
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.787533189461263e-06,
+      "loss": 1.0777,
+      "step": 392
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.786435477351353e-06,
+      "loss": 0.8623,
+      "step": 393
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.785335063380662e-06,
+      "loss": 1.678,
+      "step": 394
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7842319488495395e-06,
+      "loss": 1.1619,
+      "step": 395
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.783126135061526e-06,
+      "loss": 0.8736,
+      "step": 396
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7820176233233525e-06,
+      "loss": 0.9965,
+      "step": 397
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.780906414944938e-06,
+      "loss": 1.3315,
+      "step": 398
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.779792511239388e-06,
+      "loss": 1.3204,
+      "step": 399
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.77867591352299e-06,
+      "loss": 1.8712,
+      "step": 400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7775566231152216e-06,
+      "loss": 1.3979,
+      "step": 401
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7764346413387356e-06,
+      "loss": 1.1211,
+      "step": 402
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.775309969519369e-06,
+      "loss": 0.7957,
+      "step": 403
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.774182608986137e-06,
+      "loss": 1.152,
+      "step": 404
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.773052561071232e-06,
+      "loss": 1.5962,
+      "step": 405
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.771919827110021e-06,
+      "loss": 0.9141,
+      "step": 406
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7707844084410444e-06,
+      "loss": 1.4446,
+      "step": 407
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.769646306406019e-06,
+      "loss": 1.1342,
+      "step": 408
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.76850552234983e-06,
+      "loss": 0.8174,
+      "step": 409
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.76736205762053e-06,
+      "loss": 1.8466,
+      "step": 410
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.766215913569342e-06,
+      "loss": 0.885,
+      "step": 411
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.765067091550655e-06,
+      "loss": 0.9354,
+      "step": 412
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.763915592922022e-06,
+      "loss": 1.4256,
+      "step": 413
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.762761419044157e-06,
+      "loss": 1.0847,
+      "step": 414
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.761604571280939e-06,
+      "loss": 1.0844,
+      "step": 415
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.760445050999405e-06,
+      "loss": 1.2039,
+      "step": 416
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.759282859569749e-06,
+      "loss": 1.2732,
+      "step": 417
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7581179983653224e-06,
+      "loss": 1.2579,
+      "step": 418
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.756950468762632e-06,
+      "loss": 0.8733,
+      "step": 419
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.755780272141338e-06,
+      "loss": 1.8547,
+      "step": 420
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.75460740988425e-06,
+      "loss": 0.8285,
+      "step": 421
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.753431883377329e-06,
+      "loss": 1.0063,
+      "step": 422
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.752253694009686e-06,
+      "loss": 1.2905,
+      "step": 423
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7510728431735765e-06,
+      "loss": 1.2867,
+      "step": 424
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.749889332264401e-06,
+      "loss": 1.15,
+      "step": 425
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.748703162680704e-06,
+      "loss": 1.2434,
+      "step": 426
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.747514335824172e-06,
+      "loss": 1.1005,
+      "step": 427
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.746322853099631e-06,
+      "loss": 0.7872,
+      "step": 428
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.745128715915047e-06,
+      "loss": 1.2451,
+      "step": 429
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.743931925681519e-06,
+      "loss": 0.8286,
+      "step": 430
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.742732483813286e-06,
+      "loss": 0.8608,
+      "step": 431
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.741530391727716e-06,
+      "loss": 0.7686,
+      "step": 432
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7403256508453125e-06,
+      "loss": 0.9078,
+      "step": 433
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.739118262589706e-06,
+      "loss": 1.4336,
+      "step": 434
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.737908228387656e-06,
+      "loss": 0.7947,
+      "step": 435
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 1.2669733762741089,
+      "eval_runtime": 53.5588,
+      "eval_samples_per_second": 14.414,
+      "eval_steps_per_second": 7.207,
+      "step": 435
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.736695549669052e-06,
+      "loss": 1.4744,
+      "step": 436
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.735480227866902e-06,
+      "loss": 1.2414,
+      "step": 437
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.734262264417345e-06,
+      "loss": 0.9221,
+      "step": 438
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.733041660759635e-06,
+      "loss": 1.1588,
+      "step": 439
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.73181841833615e-06,
+      "loss": 1.019,
+      "step": 440
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.730592538592384e-06,
+      "loss": 1.3191,
+      "step": 441
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.72936402297695e-06,
+      "loss": 0.9169,
+      "step": 442
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.728132872941572e-06,
+      "loss": 0.9911,
+      "step": 443
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.726899089941091e-06,
+      "loss": 1.1003,
+      "step": 444
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.725662675433455e-06,
+      "loss": 1.4743,
+      "step": 445
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.724423630879727e-06,
+      "loss": 0.9946,
+      "step": 446
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.723181957744075e-06,
+      "loss": 1.0176,
+      "step": 447
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.721937657493771e-06,
+      "loss": 1.0483,
+      "step": 448
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.720690731599195e-06,
+      "loss": 1.6443,
+      "step": 449
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.71944118153383e-06,
+      "loss": 1.2673,
+      "step": 450
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.718189008774256e-06,
+      "loss": 1.1349,
+      "step": 451
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.716934214800155e-06,
+      "loss": 1.5773,
+      "step": 452
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7156768010943066e-06,
+      "loss": 1.4548,
+      "step": 453
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.714416769142586e-06,
+      "loss": 1.1942,
+      "step": 454
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.71315412043396e-06,
+      "loss": 1.4373,
+      "step": 455
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.711888856460493e-06,
+      "loss": 1.5013,
+      "step": 456
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.710620978717333e-06,
+      "loss": 1.3288,
+      "step": 457
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.709350488702721e-06,
+      "loss": 0.9713,
+      "step": 458
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.708077387917984e-06,
+      "loss": 1.3237,
+      "step": 459
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.706801677867535e-06,
+      "loss": 1.3598,
+      "step": 460
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.705523360058868e-06,
+      "loss": 1.0094,
+      "step": 461
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7042424360025606e-06,
+      "loss": 1.5585,
+      "step": 462
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.70295890721227e-06,
+      "loss": 1.5537,
+      "step": 463
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.70167277520473e-06,
+      "loss": 0.9304,
+      "step": 464
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.700384041499752e-06,
+      "loss": 0.9573,
+      "step": 465
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.699092707620222e-06,
+      "loss": 1.66,
+      "step": 466
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.697798775092096e-06,
+      "loss": 0.9916,
+      "step": 467
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.6965022454444046e-06,
+      "loss": 1.482,
+      "step": 468
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.695203120209245e-06,
+      "loss": 1.0236,
+      "step": 469
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.693901400921782e-06,
+      "loss": 1.093,
+      "step": 470
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.692597089120246e-06,
+      "loss": 1.2824,
+      "step": 471
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.691290186345931e-06,
+      "loss": 1.7949,
+      "step": 472
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.689980694143192e-06,
+      "loss": 1.4492,
+      "step": 473
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.688668614059445e-06,
+      "loss": 1.6656,
+      "step": 474
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.687353947645162e-06,
+      "loss": 1.5922,
+      "step": 475
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6860366964538765e-06,
+      "loss": 1.1517,
+      "step": 476
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6847168620421685e-06,
+      "loss": 1.719,
+      "step": 477
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6833944459696765e-06,
+      "loss": 1.0782,
+      "step": 478
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.682069449799089e-06,
+      "loss": 1.3512,
+      "step": 479
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.680741875096141e-06,
+      "loss": 1.9916,
+      "step": 480
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.679411723429617e-06,
+      "loss": 1.7266,
+      "step": 481
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.678078996371344e-06,
+      "loss": 1.1965,
+      "step": 482
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6767436954961956e-06,
+      "loss": 1.8546,
+      "step": 483
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.675405822382083e-06,
+      "loss": 0.9484,
+      "step": 484
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.674065378609962e-06,
+      "loss": 1.787,
+      "step": 485
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.672722365763821e-06,
+      "loss": 1.5217,
+      "step": 486
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.671376785430688e-06,
+      "loss": 0.8603,
+      "step": 487
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.670028639200622e-06,
+      "loss": 0.9619,
+      "step": 488
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.668677928666717e-06,
+      "loss": 1.5593,
+      "step": 489
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.667324655425095e-06,
+      "loss": 0.9245,
+      "step": 490
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.665968821074907e-06,
+      "loss": 0.8437,
+      "step": 491
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.66461042721833e-06,
+      "loss": 1.0066,
+      "step": 492
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.663249475460569e-06,
+      "loss": 0.8552,
+      "step": 493
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.661885967409846e-06,
+      "loss": 0.7843,
+      "step": 494
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.660519904677407e-06,
+      "loss": 1.0128,
+      "step": 495
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.659151288877517e-06,
+      "loss": 1.1489,
+      "step": 496
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.6577801216274565e-06,
+      "loss": 1.3491,
+      "step": 497
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.656406404547523e-06,
+      "loss": 0.8747,
+      "step": 498
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.6550301392610256e-06,
+      "loss": 0.9925,
+      "step": 499
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.653651327394284e-06,
+      "loss": 1.6417,
+      "step": 500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.652269970576628e-06,
+      "loss": 1.7158,
+      "step": 501
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6508860704403965e-06,
+      "loss": 1.4892,
+      "step": 502
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.649499628620931e-06,
+      "loss": 0.9178,
+      "step": 503
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.648110646756579e-06,
+      "loss": 0.8608,
+      "step": 504
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6467191264886855e-06,
+      "loss": 0.6383,
+      "step": 505
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6453250694616e-06,
+      "loss": 1.1731,
+      "step": 506
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.643928477322667e-06,
+      "loss": 1.1921,
+      "step": 507
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.642529351722227e-06,
+      "loss": 1.4647,
+      "step": 508
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.641127694313615e-06,
+      "loss": 1.8037,
+      "step": 509
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6397235067531564e-06,
+      "loss": 1.3182,
+      "step": 510
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.638316790700168e-06,
+      "loss": 1.6523,
+      "step": 511
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.636907547816953e-06,
+      "loss": 1.6149,
+      "step": 512
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.635495779768803e-06,
+      "loss": 0.9444,
+      "step": 513
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.63408148822399e-06,
+      "loss": 0.9974,
+      "step": 514
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.632664674853772e-06,
+      "loss": 0.9567,
+      "step": 515
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.631245341332383e-06,
+      "loss": 1.2978,
+      "step": 516
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.629823489337039e-06,
+      "loss": 1.3438,
+      "step": 517
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.628399120547928e-06,
+      "loss": 1.3149,
+      "step": 518
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.626972236648214e-06,
+      "loss": 1.098,
+      "step": 519
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.625542839324036e-06,
+      "loss": 1.7058,
+      "step": 520
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.624110930264498e-06,
+      "loss": 0.9291,
+      "step": 521
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.622676511161675e-06,
+      "loss": 0.846,
+      "step": 522
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6212395837106065e-06,
+      "loss": 1.5026,
+      "step": 523
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.619800149609298e-06,
+      "loss": 0.9427,
+      "step": 524
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.618358210558717e-06,
+      "loss": 0.7821,
+      "step": 525
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.616913768262789e-06,
+      "loss": 1.4015,
+      "step": 526
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6154668244284e-06,
+      "loss": 1.6013,
+      "step": 527
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.614017380765389e-06,
+      "loss": 1.6039,
+      "step": 528
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.612565438986553e-06,
+      "loss": 0.8531,
+      "step": 529
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.611111000807637e-06,
+      "loss": 1.149,
+      "step": 530
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.60965406794734e-06,
+      "loss": 1.6213,
+      "step": 531
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.608194642127304e-06,
+      "loss": 0.8759,
+      "step": 532
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.606732725072122e-06,
+      "loss": 1.0741,
+      "step": 533
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.605268318509327e-06,
+      "loss": 1.4488,
+      "step": 534
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.603801424169397e-06,
+      "loss": 1.1136,
+      "step": 535
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.602332043785748e-06,
+      "loss": 1.0564,
+      "step": 536
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.600860179094732e-06,
+      "loss": 1.7812,
+      "step": 537
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.599385831835641e-06,
+      "loss": 1.2911,
+      "step": 538
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.597909003750697e-06,
+      "loss": 1.0261,
+      "step": 539
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.596429696585054e-06,
+      "loss": 0.9223,
+      "step": 540
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.594947912086799e-06,
+      "loss": 0.936,
+      "step": 541
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.593463652006941e-06,
+      "loss": 0.8679,
+      "step": 542
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.5919769180994185e-06,
+      "loss": 1.05,
+      "step": 543
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.590487712121092e-06,
+      "loss": 0.8358,
+      "step": 544
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.588996035831742e-06,
+      "loss": 1.4855,
+      "step": 545
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.587501890994071e-06,
+      "loss": 1.3494,
+      "step": 546
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.586005279373696e-06,
+      "loss": 1.1508,
+      "step": 547
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.58450620273915e-06,
+      "loss": 0.9384,
+      "step": 548
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.583004662861877e-06,
+      "loss": 1.3447,
+      "step": 549
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.581500661516237e-06,
+      "loss": 1.4123,
+      "step": 550
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.579994200479493e-06,
+      "loss": 1.1999,
+      "step": 551
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.578485281531817e-06,
+      "loss": 1.0577,
+      "step": 552
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.576973906456286e-06,
+      "loss": 0.7701,
+      "step": 553
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.575460077038877e-06,
+      "loss": 1.0168,
+      "step": 554
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.573943795068471e-06,
+      "loss": 1.0581,
+      "step": 555
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.572425062336844e-06,
+      "loss": 1.2371,
+      "step": 556
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.570903880638669e-06,
+      "loss": 1.027,
+      "step": 557
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.569380251771514e-06,
+      "loss": 1.555,
+      "step": 558
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.567854177535837e-06,
+      "loss": 1.5865,
+      "step": 559
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.566325659734987e-06,
+      "loss": 1.59,
+      "step": 560
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.564794700175202e-06,
+      "loss": 0.7054,
+      "step": 561
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.563261300665599e-06,
+      "loss": 1.8609,
+      "step": 562
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.561725463018186e-06,
+      "loss": 1.3677,
+      "step": 563
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.560187189047849e-06,
+      "loss": 1.3487,
+      "step": 564
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.558646480572352e-06,
+      "loss": 1.2953,
+      "step": 565
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.557103339412337e-06,
+      "loss": 1.4691,
+      "step": 566
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.555557767391319e-06,
+      "loss": 1.4562,
+      "step": 567
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.5540097663356885e-06,
+      "loss": 1.0457,
+      "step": 568
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.5524593380747055e-06,
+      "loss": 0.9963,
+      "step": 569
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.550906484440495e-06,
+      "loss": 1.1226,
+      "step": 570
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.5493512072680535e-06,
+      "loss": 1.4231,
+      "step": 571
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.547793508395235e-06,
+      "loss": 1.1163,
+      "step": 572
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.546233389662763e-06,
+      "loss": 0.9493,
+      "step": 573
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5446708529142116e-06,
+      "loss": 1.4465,
+      "step": 574
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.54310589999602e-06,
+      "loss": 0.8873,
+      "step": 575
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.541538532757478e-06,
+      "loss": 1.456,
+      "step": 576
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.53996875305073e-06,
+      "loss": 0.9274,
+      "step": 577
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5383965627307705e-06,
+      "loss": 1.3981,
+      "step": 578
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.536821963655443e-06,
+      "loss": 1.3882,
+      "step": 579
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.535244957685438e-06,
+      "loss": 1.3814,
+      "step": 580
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.2516705989837646,
+      "eval_runtime": 53.6602,
+      "eval_samples_per_second": 14.387,
+      "eval_steps_per_second": 7.193,
+      "step": 580
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5336655466842904e-06,
+      "loss": 1.1422,
+      "step": 581
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.532083732518375e-06,
+      "loss": 1.042,
+      "step": 582
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.530499517056909e-06,
+      "loss": 1.1034,
+      "step": 583
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.528912902171944e-06,
+      "loss": 0.7851,
+      "step": 584
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.52732388973837e-06,
+      "loss": 1.4648,
+      "step": 585
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.52573248163391e-06,
+      "loss": 0.8926,
+      "step": 586
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.524138679739115e-06,
+      "loss": 1.1289,
+      "step": 587
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 1.074,
+      "step": 588
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.520943902114878e-06,
+      "loss": 1.179,
+      "step": 589
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.519342930160678e-06,
+      "loss": 1.0436,
+      "step": 590
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.517739571966618e-06,
+      "loss": 0.9787,
+      "step": 591
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.516133829427377e-06,
+      "loss": 1.4049,
+      "step": 592
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.514525704440446e-06,
+      "loss": 1.0072,
+      "step": 593
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.512915198906131e-06,
+      "loss": 0.9599,
+      "step": 594
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.511302314727552e-06,
+      "loss": 0.9298,
+      "step": 595
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.509687053810643e-06,
+      "loss": 0.885,
+      "step": 596
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.508069418064141e-06,
+      "loss": 0.7679,
+      "step": 597
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.506449409399592e-06,
+      "loss": 1.0181,
+      "step": 598
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.504827029731347e-06,
+      "loss": 1.3923,
+      "step": 599
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.503202280976557e-06,
+      "loss": 1.4239,
+      "step": 600
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5015751650551756e-06,
+      "loss": 1.1096,
+      "step": 601
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.499945683889948e-06,
+      "loss": 1.6356,
+      "step": 602
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4983138394064205e-06,
+      "loss": 1.4299,
+      "step": 603
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.496679633532929e-06,
+      "loss": 0.624,
+      "step": 604
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4950430682005995e-06,
+      "loss": 1.2829,
+      "step": 605
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.493404145343347e-06,
+      "loss": 1.2277,
+      "step": 606
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.491762866897872e-06,
+      "loss": 1.4662,
+      "step": 607
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.49011923480366e-06,
+      "loss": 1.0535,
+      "step": 608
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.488473251002976e-06,
+      "loss": 1.3935,
+      "step": 609
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.486824917440865e-06,
+      "loss": 0.6957,
+      "step": 610
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.485174236065149e-06,
+      "loss": 1.4466,
+      "step": 611
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.483521208826421e-06,
+      "loss": 0.655,
+      "step": 612
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4818658376780514e-06,
+      "loss": 1.1986,
+      "step": 613
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4802081245761765e-06,
+      "loss": 1.8627,
+      "step": 614
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.478548071479702e-06,
+      "loss": 0.998,
+      "step": 615
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.476885680350298e-06,
+      "loss": 1.7648,
+      "step": 616
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.475220953152397e-06,
+      "loss": 1.2204,
+      "step": 617
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.473553891853194e-06,
+      "loss": 1.2936,
+      "step": 618
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.471884498422639e-06,
+      "loss": 1.1346,
+      "step": 619
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4702127748334406e-06,
+      "loss": 1.2736,
+      "step": 620
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.46853872306106e-06,
+      "loss": 1.1092,
+      "step": 621
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4668623450837085e-06,
+      "loss": 1.0157,
+      "step": 622
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.465183642882348e-06,
+      "loss": 1.3962,
+      "step": 623
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.463502618440687e-06,
+      "loss": 1.6577,
+      "step": 624
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.461819273745176e-06,
+      "loss": 1.1963,
+      "step": 625
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.460133610785009e-06,
+      "loss": 0.9,
+      "step": 626
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.458445631552119e-06,
+      "loss": 0.961,
+      "step": 627
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.456755338041178e-06,
+      "loss": 1.7704,
+      "step": 628
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.455062732249589e-06,
+      "loss": 1.36,
+      "step": 629
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4533678161774895e-06,
+      "loss": 0.8109,
+      "step": 630
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.451670591827749e-06,
+      "loss": 1.3852,
+      "step": 631
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.44997106120596e-06,
+      "loss": 0.8871,
+      "step": 632
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4482692263204455e-06,
+      "loss": 1.3472,
+      "step": 633
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4465650891822466e-06,
+      "loss": 1.0071,
+      "step": 634
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.444858651805128e-06,
+      "loss": 1.2652,
+      "step": 635
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4431499162055724e-06,
+      "loss": 1.4467,
+      "step": 636
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.441438884402777e-06,
+      "loss": 1.472,
+      "step": 637
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4397255584186534e-06,
+      "loss": 1.4468,
+      "step": 638
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.438009940277825e-06,
+      "loss": 1.1595,
+      "step": 639
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.436292032007621e-06,
+      "loss": 1.4476,
+      "step": 640
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4345718356380795e-06,
+      "loss": 1.299,
+      "step": 641
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.432849353201941e-06,
+      "loss": 0.6383,
+      "step": 642
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.431124586734648e-06,
+      "loss": 1.1705,
+      "step": 643
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.429397538274342e-06,
+      "loss": 1.2946,
+      "step": 644
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4276682098618604e-06,
+      "loss": 0.9894,
+      "step": 645
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4259366035407375e-06,
+      "loss": 0.8213,
+      "step": 646
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4242027213571954e-06,
+      "loss": 1.7667,
+      "step": 647
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4224665653601476e-06,
+      "loss": 0.7915,
+      "step": 648
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.420728137601196e-06,
+      "loss": 1.249,
+      "step": 649
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.418987440134624e-06,
+      "loss": 1.3703,
+      "step": 650
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.417244475017398e-06,
+      "loss": 1.3207,
+      "step": 651
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.415499244309166e-06,
+      "loss": 1.793,
+      "step": 652
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4137517500722516e-06,
+      "loss": 1.4352,
+      "step": 653
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.412001994371653e-06,
+      "loss": 1.8998,
+      "step": 654
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.410249979275042e-06,
+      "loss": 1.3503,
+      "step": 655
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.408495706852758e-06,
+      "loss": 0.7287,
+      "step": 656
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4067391791778104e-06,
+      "loss": 1.3658,
+      "step": 657
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.404980398325873e-06,
+      "loss": 0.9159,
+      "step": 658
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.40321936637528e-06,
+      "loss": 1.7305,
+      "step": 659
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.40145608540703e-06,
+      "loss": 0.8456,
+      "step": 660
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.399690557504774e-06,
+      "loss": 0.8557,
+      "step": 661
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.397922784754824e-06,
+      "loss": 0.8167,
+      "step": 662
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.396152769246137e-06,
+      "loss": 1.002,
+      "step": 663
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.39438051307033e-06,
+      "loss": 1.3232,
+      "step": 664
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.392606018321659e-06,
+      "loss": 1.5761,
+      "step": 665
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.390829287097029e-06,
+      "loss": 1.1345,
+      "step": 666
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.389050321495991e-06,
+      "loss": 1.6143,
+      "step": 667
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.38726912362073e-06,
+      "loss": 0.9397,
+      "step": 668
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.385485695576071e-06,
+      "loss": 1.3128,
+      "step": 669
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.38370003946948e-06,
+      "loss": 1.3883,
+      "step": 670
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.381912157411046e-06,
+      "loss": 0.8553,
+      "step": 671
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.380122051513497e-06,
+      "loss": 1.1036,
+      "step": 672
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.378329723892184e-06,
+      "loss": 1.2276,
+      "step": 673
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.376535176665085e-06,
+      "loss": 1.4422,
+      "step": 674
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.374738411952801e-06,
+      "loss": 1.1659,
+      "step": 675
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.372939431878552e-06,
+      "loss": 1.5647,
+      "step": 676
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.371138238568179e-06,
+      "loss": 1.3946,
+      "step": 677
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.369334834150136e-06,
+      "loss": 0.7131,
+      "step": 678
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.367529220755489e-06,
+      "loss": 1.0276,
+      "step": 679
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.365721400517916e-06,
+      "loss": 1.437,
+      "step": 680
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.363911375573702e-06,
+      "loss": 0.9089,
+      "step": 681
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.362099148061739e-06,
+      "loss": 1.0697,
+      "step": 682
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.360284720123519e-06,
+      "loss": 1.4068,
+      "step": 683
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.358468093903136e-06,
+      "loss": 0.8719,
+      "step": 684
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.356649271547282e-06,
+      "loss": 0.9841,
+      "step": 685
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.354828255205242e-06,
+      "loss": 1.541,
+      "step": 686
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3530050470288955e-06,
+      "loss": 1.0059,
+      "step": 687
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3511796491727135e-06,
+      "loss": 1.1743,
+      "step": 688
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.34935206379375e-06,
+      "loss": 1.5674,
+      "step": 689
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3475222930516484e-06,
+      "loss": 1.3795,
+      "step": 690
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3456903391086315e-06,
+      "loss": 0.824,
+      "step": 691
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3438562041295045e-06,
+      "loss": 0.8724,
+      "step": 692
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.342019890281648e-06,
+      "loss": 1.026,
+      "step": 693
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.340181399735018e-06,
+      "loss": 0.9268,
+      "step": 694
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.338340734662143e-06,
+      "loss": 1.4699,
+      "step": 695
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.33649789723812e-06,
+      "loss": 1.5072,
+      "step": 696
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.334652889640615e-06,
+      "loss": 1.0228,
+      "step": 697
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.332805714049856e-06,
+      "loss": 0.9188,
+      "step": 698
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.330956372648635e-06,
+      "loss": 1.0169,
+      "step": 699
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.329104867622302e-06,
+      "loss": 0.7205,
+      "step": 700
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3272512011587665e-06,
+      "loss": 1.1265,
+      "step": 701
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.325395375448487e-06,
+      "loss": 1.3212,
+      "step": 702
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.323537392684477e-06,
+      "loss": 0.8664,
+      "step": 703
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3216772550623e-06,
+      "loss": 1.3096,
+      "step": 704
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.319814964780063e-06,
+      "loss": 1.6955,
+      "step": 705
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.317950524038419e-06,
+      "loss": 1.4587,
+      "step": 706
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.316083935040561e-06,
+      "loss": 1.2825,
+      "step": 707
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.31421519999222e-06,
+      "loss": 1.1505,
+      "step": 708
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3123443211016655e-06,
+      "loss": 1.1944,
+      "step": 709
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.310471300579697e-06,
+      "loss": 1.0146,
+      "step": 710
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.308596140639647e-06,
+      "loss": 1.1227,
+      "step": 711
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3067188434973765e-06,
+      "loss": 1.0811,
+      "step": 712
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3048394113712696e-06,
+      "loss": 0.9842,
+      "step": 713
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.302957846482236e-06,
+      "loss": 1.7324,
+      "step": 714
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3010741510537046e-06,
+      "loss": 1.1923,
+      "step": 715
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.299188327311621e-06,
+      "loss": 1.1261,
+      "step": 716
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.2973003774844464e-06,
+      "loss": 1.5921,
+      "step": 717
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.295410303803155e-06,
+      "loss": 0.9543,
+      "step": 718
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2935181085012306e-06,
+      "loss": 1.0813,
+      "step": 719
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2916237938146635e-06,
+      "loss": 1.7461,
+      "step": 720
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2897273619819494e-06,
+      "loss": 1.2278,
+      "step": 721
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.287828815244083e-06,
+      "loss": 0.8105,
+      "step": 722
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2859281558445635e-06,
+      "loss": 0.8013,
+      "step": 723
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.284025386029381e-06,
+      "loss": 0.977,
+      "step": 724
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.282120508047023e-06,
+      "loss": 1.7712,
+      "step": 725
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.2422988414764404,
+      "eval_runtime": 53.6577,
+      "eval_samples_per_second": 14.387,
+      "eval_steps_per_second": 7.194,
+      "step": 725
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2802135241484676e-06,
+      "loss": 0.9182,
+      "step": 726
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.27830443658718e-06,
+      "loss": 0.9902,
+      "step": 727
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2763932476191125e-06,
+      "loss": 1.3087,
+      "step": 728
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.274479959502701e-06,
+      "loss": 0.7967,
+      "step": 729
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.27256457449886e-06,
+      "loss": 1.2144,
+      "step": 730
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.270647094870985e-06,
+      "loss": 0.8462,
+      "step": 731
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.268727522884942e-06,
+      "loss": 1.4271,
+      "step": 732
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.266805860809076e-06,
+      "loss": 1.2957,
+      "step": 733
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.264882110914196e-06,
+      "loss": 1.1985,
+      "step": 734
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.262956275473581e-06,
+      "loss": 1.2892,
+      "step": 735
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.261028356762973e-06,
+      "loss": 1.0717,
+      "step": 736
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.259098357060578e-06,
+      "loss": 0.948,
+      "step": 737
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.25716627864706e-06,
+      "loss": 0.9578,
+      "step": 738
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.255232123805537e-06,
+      "loss": 1.4776,
+      "step": 739
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.253295894821585e-06,
+      "loss": 0.6842,
+      "step": 740
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.251357593983228e-06,
+      "loss": 1.3206,
+      "step": 741
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.249417223580939e-06,
+      "loss": 1.1884,
+      "step": 742
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.247474785907636e-06,
+      "loss": 1.0852,
+      "step": 743
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.245530283258682e-06,
+      "loss": 0.9386,
+      "step": 744
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.24358371793188e-06,
+      "loss": 1.8676,
+      "step": 745
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.241635092227465e-06,
+      "loss": 0.8333,
+      "step": 746
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.239684408448114e-06,
+      "loss": 1.4695,
+      "step": 747
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.2377316688989326e-06,
+      "loss": 1.4006,
+      "step": 748
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.235776875887455e-06,
+      "loss": 1.5704,
+      "step": 749
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.233820031723644e-06,
+      "loss": 0.989,
+      "step": 750
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.231861138719886e-06,
+      "loss": 0.8683,
+      "step": 751
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.229900199190985e-06,
+      "loss": 1.1916,
+      "step": 752
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.227937215454169e-06,
+      "loss": 1.6889,
+      "step": 753
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.225972189829076e-06,
+      "loss": 0.9639,
+      "step": 754
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.224005124637762e-06,
+      "loss": 1.4438,
+      "step": 755
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.222036022204689e-06,
+      "loss": 0.6203,
+      "step": 756
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.220064884856729e-06,
+      "loss": 0.8918,
+      "step": 757
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.218091714923157e-06,
+      "loss": 0.9794,
+      "step": 758
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.216116514735652e-06,
+      "loss": 1.5727,
+      "step": 759
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.21413928662829e-06,
+      "loss": 1.9055,
+      "step": 760
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.212160032937543e-06,
+      "loss": 1.7443,
+      "step": 761
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.210178756002279e-06,
+      "loss": 1.2116,
+      "step": 762
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.208195458163755e-06,
+      "loss": 0.892,
+      "step": 763
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.206210141765618e-06,
+      "loss": 1.3735,
+      "step": 764
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.204222809153897e-06,
+      "loss": 1.1583,
+      "step": 765
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.202233462677008e-06,
+      "loss": 1.0731,
+      "step": 766
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.200242104685742e-06,
+      "loss": 0.9748,
+      "step": 767
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.198248737533271e-06,
+      "loss": 1.2398,
+      "step": 768
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1962533635751365e-06,
+      "loss": 1.5763,
+      "step": 769
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.194255985169259e-06,
+      "loss": 0.7805,
+      "step": 770
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.19225660467592e-06,
+      "loss": 1.4626,
+      "step": 771
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.19025522445777e-06,
+      "loss": 0.7902,
+      "step": 772
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1882518468798235e-06,
+      "loss": 1.3892,
+      "step": 773
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.186246474309455e-06,
+      "loss": 0.9948,
+      "step": 774
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.184239109116393e-06,
+      "loss": 1.5633,
+      "step": 775
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.182229753672726e-06,
+      "loss": 0.8679,
+      "step": 776
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.18021841035289e-06,
+      "loss": 0.7651,
+      "step": 777
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1782050815336725e-06,
+      "loss": 1.3411,
+      "step": 778
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1761897695942076e-06,
+      "loss": 1.6998,
+      "step": 779
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.17417247691597e-06,
+      "loss": 1.1193,
+      "step": 780
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.172153205882777e-06,
+      "loss": 1.437,
+      "step": 781
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.170131958880784e-06,
+      "loss": 1.0518,
+      "step": 782
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.168108738298481e-06,
+      "loss": 0.9892,
+      "step": 783
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.166083546526689e-06,
+      "loss": 1.4121,
+      "step": 784
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.164056385958561e-06,
+      "loss": 1.8558,
+      "step": 785
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.162027258989575e-06,
+      "loss": 1.6905,
+      "step": 786
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.159996168017531e-06,
+      "loss": 1.3669,
+      "step": 787
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.157963115442552e-06,
+      "loss": 0.9603,
+      "step": 788
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1559281036670785e-06,
+      "loss": 0.7815,
+      "step": 789
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.153891135095866e-06,
+      "loss": 1.6807,
+      "step": 790
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.151852212135984e-06,
+      "loss": 0.5376,
+      "step": 791
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.149811337196808e-06,
+      "loss": 1.0208,
+      "step": 792
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1477685126900215e-06,
+      "loss": 1.3474,
+      "step": 793
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.145723741029613e-06,
+      "loss": 1.1992,
+      "step": 794
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1436770246318725e-06,
+      "loss": 1.1475,
+      "step": 795
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.141628365915386e-06,
+      "loss": 1.0059,
+      "step": 796
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.139577767301035e-06,
+      "loss": 0.8612,
+      "step": 797
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.137525231211995e-06,
+      "loss": 1.7248,
+      "step": 798
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1354707600737286e-06,
+      "loss": 1.5388,
+      "step": 799
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.133414356313988e-06,
+      "loss": 1.2048,
+      "step": 800
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.131356022362807e-06,
+      "loss": 0.9788,
+      "step": 801
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.129295760652499e-06,
+      "loss": 1.2049,
+      "step": 802
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.12723357361766e-06,
+      "loss": 1.7343,
+      "step": 803
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.125169463695156e-06,
+      "loss": 0.7971,
+      "step": 804
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1231034333241295e-06,
+      "loss": 0.8876,
+      "step": 805
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.12103548494599e-06,
+      "loss": 1.6504,
+      "step": 806
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.118965621004413e-06,
+      "loss": 1.1006,
+      "step": 807
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.116893843945339e-06,
+      "loss": 0.9648,
+      "step": 808
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.114820156216969e-06,
+      "loss": 0.859,
+      "step": 809
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.112744560269762e-06,
+      "loss": 1.0627,
+      "step": 810
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.11066705855643e-06,
+      "loss": 1.1119,
+      "step": 811
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.108587653531942e-06,
+      "loss": 1.1377,
+      "step": 812
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.106506347653509e-06,
+      "loss": 0.7878,
+      "step": 813
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.104423143380594e-06,
+      "loss": 1.485,
+      "step": 814
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1023380431749e-06,
+      "loss": 1.3263,
+      "step": 815
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.100251049500373e-06,
+      "loss": 1.4766,
+      "step": 816
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.098162164823194e-06,
+      "loss": 1.094,
+      "step": 817
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.096071391611781e-06,
+      "loss": 0.9679,
+      "step": 818
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.093978732336781e-06,
+      "loss": 0.9718,
+      "step": 819
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.091884189471071e-06,
+      "loss": 1.0933,
+      "step": 820
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.089787765489757e-06,
+      "loss": 1.2173,
+      "step": 821
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.087689462870161e-06,
+      "loss": 1.3192,
+      "step": 822
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.08558928409183e-06,
+      "loss": 1.3672,
+      "step": 823
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.083487231636528e-06,
+      "loss": 0.9721,
+      "step": 824
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.081383307988232e-06,
+      "loss": 0.8727,
+      "step": 825
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.079277515633127e-06,
+      "loss": 1.194,
+      "step": 826
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.077169857059614e-06,
+      "loss": 0.7615,
+      "step": 827
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.0750603347582915e-06,
+      "loss": 1.2318,
+      "step": 828
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.072948951221965e-06,
+      "loss": 0.8819,
+      "step": 829
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.070835708945637e-06,
+      "loss": 1.4115,
+      "step": 830
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.068720610426508e-06,
+      "loss": 0.7792,
+      "step": 831
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.066603658163972e-06,
+      "loss": 0.8577,
+      "step": 832
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.064484854659611e-06,
+      "loss": 1.3564,
+      "step": 833
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.062364202417198e-06,
+      "loss": 1.0379,
+      "step": 834
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.060241703942689e-06,
+      "loss": 1.089,
+      "step": 835
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.058117361744222e-06,
+      "loss": 0.9309,
+      "step": 836
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.055991178332113e-06,
+      "loss": 1.0059,
+      "step": 837
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.053863156218855e-06,
+      "loss": 0.98,
+      "step": 838
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0517332979191126e-06,
+      "loss": 1.2531,
+      "step": 839
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.04960160594972e-06,
+      "loss": 1.6179,
+      "step": 840
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0474680828296805e-06,
+      "loss": 0.9697,
+      "step": 841
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.045332731080159e-06,
+      "loss": 1.9022,
+      "step": 842
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.043195553224482e-06,
+      "loss": 1.8396,
+      "step": 843
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.041056551788134e-06,
+      "loss": 1.2137,
+      "step": 844
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.038915729298755e-06,
+      "loss": 1.0898,
+      "step": 845
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0367730882861335e-06,
+      "loss": 0.9471,
+      "step": 846
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0346286312822116e-06,
+      "loss": 0.891,
+      "step": 847
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.032482360821076e-06,
+      "loss": 1.0731,
+      "step": 848
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.030334279438954e-06,
+      "loss": 1.1644,
+      "step": 849
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.028184389674216e-06,
+      "loss": 1.1629,
+      "step": 850
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0260326940673665e-06,
+      "loss": 1.4033,
+      "step": 851
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.023879195161047e-06,
+      "loss": 0.9199,
+      "step": 852
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.021723895500027e-06,
+      "loss": 1.6712,
+      "step": 853
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.019566797631206e-06,
+      "loss": 1.7075,
+      "step": 854
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.017407904103607e-06,
+      "loss": 1.2572,
+      "step": 855
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.015247217468377e-06,
+      "loss": 1.4046,
+      "step": 856
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.013084740278778e-06,
+      "loss": 1.5426,
+      "step": 857
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.010920475090192e-06,
+      "loss": 1.4335,
+      "step": 858
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.008754424460112e-06,
+      "loss": 0.8425,
+      "step": 859
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.006586590948141e-06,
+      "loss": 1.1965,
+      "step": 860
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.004416977115989e-06,
+      "loss": 0.8491,
+      "step": 861
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.00224558552747e-06,
+      "loss": 0.9847,
+      "step": 862
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.000072418748498e-06,
+      "loss": 1.106,
+      "step": 863
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.9978974793470855e-06,
+      "loss": 1.0784,
+      "step": 864
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.9957207698933385e-06,
+      "loss": 1.0359,
+      "step": 865
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.993542292959456e-06,
+      "loss": 0.9047,
+      "step": 866
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.991362051119725e-06,
+      "loss": 0.7676,
+      "step": 867
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.989180046950518e-06,
+      "loss": 0.6939,
+      "step": 868
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.986996283030291e-06,
+      "loss": 1.286,
+      "step": 869
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.984810761939578e-06,
+      "loss": 1.2697,
+      "step": 870
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.2340365648269653,
+      "eval_runtime": 54.4194,
+      "eval_samples_per_second": 14.186,
+      "eval_steps_per_second": 7.093,
+      "step": 870
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.98262348626099e-06,
+      "loss": 1.2608,
+      "step": 871
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.980434458579213e-06,
+      "loss": 1.5733,
+      "step": 872
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.978243681481e-06,
+      "loss": 0.756,
+      "step": 873
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.976051157555174e-06,
+      "loss": 1.592,
+      "step": 874
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.973856889392622e-06,
+      "loss": 1.2678,
+      "step": 875
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.97166087958629e-06,
+      "loss": 1.2461,
+      "step": 876
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.969463130731183e-06,
+      "loss": 1.2391,
+      "step": 877
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.967263645424364e-06,
+      "loss": 1.6344,
+      "step": 878
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.965062426264944e-06,
+      "loss": 1.6889,
+      "step": 879
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.962859475854085e-06,
+      "loss": 0.9765,
+      "step": 880
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.960654796794993e-06,
+      "loss": 0.7538,
+      "step": 881
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.9584483916929195e-06,
+      "loss": 1.0218,
+      "step": 882
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.956240263155153e-06,
+      "loss": 1.0937,
+      "step": 883
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.954030413791021e-06,
+      "loss": 1.7373,
+      "step": 884
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.951818846211882e-06,
+      "loss": 0.9229,
+      "step": 885
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.949605563031126e-06,
+      "loss": 0.766,
+      "step": 886
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.9473905668641705e-06,
+      "loss": 1.7482,
+      "step": 887
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.945173860328457e-06,
+      "loss": 0.855,
+      "step": 888
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.942955446043448e-06,
+      "loss": 1.6474,
+      "step": 889
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.940735326630626e-06,
+      "loss": 1.6998,
+      "step": 890
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.938513504713485e-06,
+      "loss": 1.1006,
+      "step": 891
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.936289982917532e-06,
+      "loss": 0.9537,
+      "step": 892
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9340647638702835e-06,
+      "loss": 1.0979,
+      "step": 893
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9318378502012636e-06,
+      "loss": 1.5734,
+      "step": 894
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.929609244541994e-06,
+      "loss": 1.5724,
+      "step": 895
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9273789495259995e-06,
+      "loss": 1.1793,
+      "step": 896
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.925146967788801e-06,
+      "loss": 1.1859,
+      "step": 897
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.92291330196791e-06,
+      "loss": 1.4159,
+      "step": 898
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9206779547028305e-06,
+      "loss": 1.6399,
+      "step": 899
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9184409286350535e-06,
+      "loss": 1.4543,
+      "step": 900
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.916202226408053e-06,
+      "loss": 0.7884,
+      "step": 901
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.913961850667283e-06,
+      "loss": 1.5351,
+      "step": 902
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.911719804060177e-06,
+      "loss": 0.8836,
+      "step": 903
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9094760892361405e-06,
+      "loss": 0.8275,
+      "step": 904
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.907230708846552e-06,
+      "loss": 1.0969,
+      "step": 905
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.904983665544759e-06,
+      "loss": 1.7432,
+      "step": 906
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9027349619860726e-06,
+      "loss": 0.8615,
+      "step": 907
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.900484600827765e-06,
+      "loss": 1.4718,
+      "step": 908
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.898232584729069e-06,
+      "loss": 1.5248,
+      "step": 909
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.895978916351172e-06,
+      "loss": 1.6631,
+      "step": 910
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.893723598357214e-06,
+      "loss": 1.0259,
+      "step": 911
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.891466633412285e-06,
+      "loss": 1.6557,
+      "step": 912
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.889208024183422e-06,
+      "loss": 1.5224,
+      "step": 913
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.886947773339601e-06,
+      "loss": 0.9531,
+      "step": 914
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.884685883551742e-06,
+      "loss": 1.0861,
+      "step": 915
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.882422357492701e-06,
+      "loss": 0.8796,
+      "step": 916
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.880157197837267e-06,
+      "loss": 1.6122,
+      "step": 917
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.877890407262157e-06,
+      "loss": 1.6063,
+      "step": 918
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.875621988446019e-06,
+      "loss": 0.8907,
+      "step": 919
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.873351944069425e-06,
+      "loss": 0.834,
+      "step": 920
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.8710802768148644e-06,
+      "loss": 0.9929,
+      "step": 921
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.868806989366748e-06,
+      "loss": 1.2989,
+      "step": 922
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.8665320844114e-06,
+      "loss": 1.3995,
+      "step": 923
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.864255564637053e-06,
+      "loss": 1.8592,
+      "step": 924
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.861977432733853e-06,
+      "loss": 0.8865,
+      "step": 925
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.859697691393849e-06,
+      "loss": 1.1653,
+      "step": 926
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.857416343310989e-06,
+      "loss": 1.4607,
+      "step": 927
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.855133391181124e-06,
+      "loss": 1.2872,
+      "step": 928
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.852848837701997e-06,
+      "loss": 0.947,
+      "step": 929
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.850562685573246e-06,
+      "loss": 0.7027,
+      "step": 930
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.848274937496397e-06,
+      "loss": 0.9595,
+      "step": 931
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.845985596174861e-06,
+      "loss": 1.1452,
+      "step": 932
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.843694664313932e-06,
+      "loss": 1.1997,
+      "step": 933
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.8414021446207866e-06,
+      "loss": 1.6929,
+      "step": 934
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.839108039804473e-06,
+      "loss": 1.2888,
+      "step": 935
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.836812352575916e-06,
+      "loss": 1.4976,
+      "step": 936
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.834515085647909e-06,
+      "loss": 1.5642,
+      "step": 937
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.832216241735111e-06,
+      "loss": 1.3293,
+      "step": 938
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.8299158235540475e-06,
+      "loss": 1.3143,
+      "step": 939
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.827613833823101e-06,
+      "loss": 1.5578,
+      "step": 940
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.825310275262513e-06,
+      "loss": 0.7517,
+      "step": 941
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.823005150594379e-06,
+      "loss": 0.9068,
+      "step": 942
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.820698462542644e-06,
+      "loss": 1.0884,
+      "step": 943
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.818390213833102e-06,
+      "loss": 0.7859,
+      "step": 944
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.81608040719339e-06,
+      "loss": 1.4531,
+      "step": 945
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.8137690453529854e-06,
+      "loss": 1.4531,
+      "step": 946
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.811456131043205e-06,
+      "loss": 1.3391,
+      "step": 947
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.8091416669971994e-06,
+      "loss": 1.247,
+      "step": 948
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.8068256559499493e-06,
+      "loss": 1.1898,
+      "step": 949
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.8045081006382668e-06,
+      "loss": 1.0309,
+      "step": 950
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.8021890038007848e-06,
+      "loss": 0.9196,
+      "step": 951
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.79986836817796e-06,
+      "loss": 1.236,
+      "step": 952
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.79754619651207e-06,
+      "loss": 1.0914,
+      "step": 953
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.7952224915472003e-06,
+      "loss": 1.0782,
+      "step": 954
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.7928972560292556e-06,
+      "loss": 0.898,
+      "step": 955
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.790570492705945e-06,
+      "loss": 1.2418,
+      "step": 956
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.788242204326785e-06,
+      "loss": 0.5817,
+      "step": 957
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.785912393643094e-06,
+      "loss": 1.2435,
+      "step": 958
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.7835810634079888e-06,
+      "loss": 1.692,
+      "step": 959
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.7812482163763807e-06,
+      "loss": 1.5054,
+      "step": 960
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.7789138553049767e-06,
+      "loss": 0.7875,
+      "step": 961
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.7765779829522674e-06,
+      "loss": 1.6516,
+      "step": 962
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.774240602078536e-06,
+      "loss": 0.909,
+      "step": 963
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.771901715445845e-06,
+      "loss": 1.5424,
+      "step": 964
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.769561325818033e-06,
+      "loss": 1.1065,
+      "step": 965
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.7672194359607213e-06,
+      "loss": 1.0141,
+      "step": 966
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.7648760486412993e-06,
+      "loss": 1.173,
+      "step": 967
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.762531166628927e-06,
+      "loss": 1.3394,
+      "step": 968
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.7601847926945318e-06,
+      "loss": 1.8383,
+      "step": 969
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.757836929610803e-06,
+      "loss": 0.7631,
+      "step": 970
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.7554875801521882e-06,
+      "loss": 0.7929,
+      "step": 971
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.7531367470948953e-06,
+      "loss": 1.2148,
+      "step": 972
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.750784433216883e-06,
+      "loss": 1.717,
+      "step": 973
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.7484306412978583e-06,
+      "loss": 1.0289,
+      "step": 974
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.7460753741192777e-06,
+      "loss": 1.4954,
+      "step": 975
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.7437186344643395e-06,
+      "loss": 1.3094,
+      "step": 976
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.7413604251179826e-06,
+      "loss": 1.6131,
+      "step": 977
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.739000748866882e-06,
+      "loss": 1.0538,
+      "step": 978
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.736639608499448e-06,
+      "loss": 1.4786,
+      "step": 979
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.7342770068058165e-06,
+      "loss": 1.5938,
+      "step": 980
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.731912946577855e-06,
+      "loss": 1.3353,
+      "step": 981
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.729547430609153e-06,
+      "loss": 0.8837,
+      "step": 982
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.727180461695019e-06,
+      "loss": 1.1192,
+      "step": 983
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.724812042632481e-06,
+      "loss": 0.8178,
+      "step": 984
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.722442176220278e-06,
+      "loss": 0.7357,
+      "step": 985
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.7200708652588603e-06,
+      "loss": 1.5317,
+      "step": 986
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.7176981125503846e-06,
+      "loss": 0.7846,
+      "step": 987
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.7153239208987136e-06,
+      "loss": 0.8197,
+      "step": 988
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.7129482931094085e-06,
+      "loss": 1.4203,
+      "step": 989
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.710571231989727e-06,
+      "loss": 0.6815,
+      "step": 990
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.7081927403486217e-06,
+      "loss": 1.0855,
+      "step": 991
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.705812820996736e-06,
+      "loss": 1.5208,
+      "step": 992
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.7034314767463985e-06,
+      "loss": 1.912,
+      "step": 993
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.7010487104116242e-06,
+      "loss": 0.7749,
+      "step": 994
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6986645248081066e-06,
+      "loss": 1.3164,
+      "step": 995
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6962789227532165e-06,
+      "loss": 1.1781,
+      "step": 996
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6938919070660003e-06,
+      "loss": 1.2583,
+      "step": 997
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.691503480567171e-06,
+      "loss": 1.3813,
+      "step": 998
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6891136460791143e-06,
+      "loss": 1.2476,
+      "step": 999
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6867224064258754e-06,
+      "loss": 0.9396,
+      "step": 1000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6843297644331612e-06,
+      "loss": 0.8906,
+      "step": 1001
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6819357229283346e-06,
+      "loss": 0.8506,
+      "step": 1002
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6795402847404153e-06,
+      "loss": 0.9023,
+      "step": 1003
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.677143452700071e-06,
+      "loss": 1.518,
+      "step": 1004
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.674745229639617e-06,
+      "loss": 1.354,
+      "step": 1005
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.6723456183930128e-06,
+      "loss": 1.6563,
+      "step": 1006
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.669944621795857e-06,
+      "loss": 1.2295,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6675422426853883e-06,
+      "loss": 1.6811,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.665138483900475e-06,
+      "loss": 1.1634,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.662733348281619e-06,
+      "loss": 1.6505,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.660326838670948e-06,
+      "loss": 1.9781,
+      "step": 1011
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6579189579122115e-06,
+      "loss": 1.4078,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6555097088507837e-06,
+      "loss": 1.2795,
+      "step": 1013
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6530990943336515e-06,
+      "loss": 0.8649,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6506871172094173e-06,
+      "loss": 1.4015,
+      "step": 1015
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 1.2239829301834106,
+      "eval_runtime": 53.5468,
+      "eval_samples_per_second": 14.417,
+      "eval_steps_per_second": 7.209,
+      "step": 1015
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.648273780328293e-06,
+      "loss": 0.8306,
+      "step": 1016
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6458590865420995e-06,
+      "loss": 0.7629,
+      "step": 1017
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.643443038704257e-06,
+      "loss": 0.7034,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6410256396697895e-06,
+      "loss": 1.8372,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.638606892295315e-06,
+      "loss": 1.2091,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.6361867994390478e-06,
+      "loss": 0.8437,
+      "step": 1021
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.633765363960789e-06,
+      "loss": 0.8183,
+      "step": 1022
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6313425887219276e-06,
+      "loss": 0.9,
+      "step": 1023
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.628918476585437e-06,
+      "loss": 1.1369,
+      "step": 1024
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6264930304158697e-06,
+      "loss": 1.0595,
+      "step": 1025
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6240662530793518e-06,
+      "loss": 0.7621,
+      "step": 1026
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6216381474435867e-06,
+      "loss": 1.5716,
+      "step": 1027
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6192087163778456e-06,
+      "loss": 0.7501,
+      "step": 1028
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6167779627529654e-06,
+      "loss": 0.8721,
+      "step": 1029
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6143458894413463e-06,
+      "loss": 0.8237,
+      "step": 1030
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.611912499316948e-06,
+      "loss": 1.267,
+      "step": 1031
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6094777952552864e-06,
+      "loss": 1.206,
+      "step": 1032
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.60704178013343e-06,
+      "loss": 0.8337,
+      "step": 1033
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6046044568299973e-06,
+      "loss": 1.7589,
+      "step": 1034
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.6021658282251504e-06,
+      "loss": 0.9766,
+      "step": 1035
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.5997258972005967e-06,
+      "loss": 1.5177,
+      "step": 1036
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.59728466663958e-06,
+      "loss": 1.4695,
+      "step": 1037
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.594842139426882e-06,
+      "loss": 0.9031,
+      "step": 1038
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5923983184488167e-06,
+      "loss": 1.1727,
+      "step": 1039
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5899532065932236e-06,
+      "loss": 1.071,
+      "step": 1040
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5875068067494716e-06,
+      "loss": 1.2549,
+      "step": 1041
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5850591218084498e-06,
+      "loss": 0.7977,
+      "step": 1042
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5826101546625653e-06,
+      "loss": 0.8636,
+      "step": 1043
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.580159908205742e-06,
+      "loss": 1.8348,
+      "step": 1044
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5777083853334145e-06,
+      "loss": 1.0609,
+      "step": 1045
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.575255588942525e-06,
+      "loss": 0.9279,
+      "step": 1046
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5728015219315226e-06,
+      "loss": 1.1864,
+      "step": 1047
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5703461872003566e-06,
+      "loss": 1.1254,
+      "step": 1048
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5678895876504745e-06,
+      "loss": 1.0267,
+      "step": 1049
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.5654317261848182e-06,
+      "loss": 1.3926,
+      "step": 1050
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.562972605707821e-06,
+      "loss": 0.8434,
+      "step": 1051
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5605122291254048e-06,
+      "loss": 1.1,
+      "step": 1052
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5580505993449744e-06,
+      "loss": 0.8945,
+      "step": 1053
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.555587719275416e-06,
+      "loss": 0.8897,
+      "step": 1054
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5531235918270934e-06,
+      "loss": 1.0842,
+      "step": 1055
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5506582199118457e-06,
+      "loss": 0.919,
+      "step": 1056
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5481916064429796e-06,
+      "loss": 0.9278,
+      "step": 1057
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.545723754335272e-06,
+      "loss": 1.544,
+      "step": 1058
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5432546665049605e-06,
+      "loss": 0.7575,
+      "step": 1059
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.540784345869746e-06,
+      "loss": 1.0788,
+      "step": 1060
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.538312795348785e-06,
+      "loss": 0.7757,
+      "step": 1061
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5358400178626856e-06,
+      "loss": 1.0101,
+      "step": 1062
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.533366016333509e-06,
+      "loss": 1.2068,
+      "step": 1063
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.530890793684759e-06,
+      "loss": 0.8573,
+      "step": 1064
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5284143528413866e-06,
+      "loss": 1.2888,
+      "step": 1065
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.525936696729779e-06,
+      "loss": 1.4604,
+      "step": 1066
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.523457828277762e-06,
+      "loss": 0.877,
+      "step": 1067
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.520977750414591e-06,
+      "loss": 1.7115,
+      "step": 1068
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.518496466070954e-06,
+      "loss": 1.3532,
+      "step": 1069
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.5160139781789615e-06,
+      "loss": 1.1231,
+      "step": 1070
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.5135302896721486e-06,
+      "loss": 1.1528,
+      "step": 1071
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.5110454034854684e-06,
+      "loss": 0.8566,
+      "step": 1072
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.5085593225552887e-06,
+      "loss": 1.4349,
+      "step": 1073
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.5060720498193895e-06,
+      "loss": 1.0565,
+      "step": 1074
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.50358358821696e-06,
+      "loss": 0.8588,
+      "step": 1075
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.501093940688592e-06,
+      "loss": 1.8337,
+      "step": 1076
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.4986031101762806e-06,
+      "loss": 0.9244,
+      "step": 1077
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.4961110996234205e-06,
+      "loss": 0.6638,
+      "step": 1078
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.4936179119747958e-06,
+      "loss": 1.3721,
+      "step": 1079
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.491123550176586e-06,
+      "loss": 1.8717,
+      "step": 1080
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.4886280171763563e-06,
+      "loss": 1.4513,
+      "step": 1081
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.486131315923057e-06,
+      "loss": 1.0628,
+      "step": 1082
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.483633449367017e-06,
+      "loss": 1.0992,
+      "step": 1083
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.481134420459944e-06,
+      "loss": 0.6331,
+      "step": 1084
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.4786342321549178e-06,
+      "loss": 0.8425,
+      "step": 1085
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.47613288740639e-06,
+      "loss": 0.9034,
+      "step": 1086
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.4736303891701766e-06,
+      "loss": 1.6633,
+      "step": 1087
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.471126740403459e-06,
+      "loss": 1.4944,
+      "step": 1088
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.4686219440647774e-06,
+      "loss": 1.9144,
+      "step": 1089
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.4661160031140257e-06,
+      "loss": 0.8263,
+      "step": 1090
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.463608920512454e-06,
+      "loss": 0.8001,
+      "step": 1091
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.4611006992226594e-06,
+      "loss": 0.8235,
+      "step": 1092
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.4585913422085848e-06,
+      "loss": 0.9863,
+      "step": 1093
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.4560808524355162e-06,
+      "loss": 0.7479,
+      "step": 1094
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.453569232870077e-06,
+      "loss": 0.9936,
+      "step": 1095
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4510564864802255e-06,
+      "loss": 1.0067,
+      "step": 1096
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4485426162352525e-06,
+      "loss": 0.7327,
+      "step": 1097
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.446027625105776e-06,
+      "loss": 1.2916,
+      "step": 1098
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.44351151606374e-06,
+      "loss": 1.382,
+      "step": 1099
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.440994292082408e-06,
+      "loss": 0.936,
+      "step": 1100
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.438475956136361e-06,
+      "loss": 0.7885,
+      "step": 1101
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4359565112014957e-06,
+      "loss": 0.7974,
+      "step": 1102
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.433435960255017e-06,
+      "loss": 1.2391,
+      "step": 1103
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.430914306275438e-06,
+      "loss": 0.7443,
+      "step": 1104
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4283915522425758e-06,
+      "loss": 1.525,
+      "step": 1105
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4258677011375473e-06,
+      "loss": 0.9756,
+      "step": 1106
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4233427559427634e-06,
+      "loss": 1.2988,
+      "step": 1107
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4208167196419312e-06,
+      "loss": 1.3856,
+      "step": 1108
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.4182895952200448e-06,
+      "loss": 1.0378,
+      "step": 1109
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.4157613856633847e-06,
+      "loss": 1.2839,
+      "step": 1110
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.4132320939595156e-06,
+      "loss": 0.8511,
+      "step": 1111
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.4107017230972777e-06,
+      "loss": 0.8881,
+      "step": 1112
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.4081702760667894e-06,
+      "loss": 0.9388,
+      "step": 1113
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.405637755859439e-06,
+      "loss": 1.3218,
+      "step": 1114
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.403104165467883e-06,
+      "loss": 1.2651,
+      "step": 1115
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.4005695078860428e-06,
+      "loss": 1.6755,
+      "step": 1116
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.3980337861091028e-06,
+      "loss": 1.5393,
+      "step": 1117
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.395497003133501e-06,
+      "loss": 1.2368,
+      "step": 1118
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.392959161956933e-06,
+      "loss": 1.2073,
+      "step": 1119
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.3904202655783426e-06,
+      "loss": 1.2589,
+      "step": 1120
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.3878803169979225e-06,
+      "loss": 0.9366,
+      "step": 1121
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.385339319217107e-06,
+      "loss": 1.101,
+      "step": 1122
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.3827972752385706e-06,
+      "loss": 1.3452,
+      "step": 1123
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.380254188066226e-06,
+      "loss": 1.3305,
+      "step": 1124
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.377710060705215e-06,
+      "loss": 1.3394,
+      "step": 1125
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.375164896161912e-06,
+      "loss": 0.9721,
+      "step": 1126
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.3726186974439144e-06,
+      "loss": 1.0626,
+      "step": 1127
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.3700714675600443e-06,
+      "loss": 1.1734,
+      "step": 1128
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.36752320952034e-06,
+      "loss": 1.2122,
+      "step": 1129
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.364973926336057e-06,
+      "loss": 1.5386,
+      "step": 1130
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.3624236210196594e-06,
+      "loss": 1.0616,
+      "step": 1131
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.359872296584821e-06,
+      "loss": 1.0179,
+      "step": 1132
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.35731995604642e-06,
+      "loss": 0.9667,
+      "step": 1133
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.354766602420534e-06,
+      "loss": 0.9075,
+      "step": 1134
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.3522122387244393e-06,
+      "loss": 1.8509,
+      "step": 1135
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.3496568679766068e-06,
+      "loss": 1.3092,
+      "step": 1136
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.347100493196692e-06,
+      "loss": 0.8243,
+      "step": 1137
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.3445431174055436e-06,
+      "loss": 1.0863,
+      "step": 1138
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3419847436251897e-06,
+      "loss": 1.0101,
+      "step": 1139
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.339425374878836e-06,
+      "loss": 0.8136,
+      "step": 1140
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3368650141908693e-06,
+      "loss": 0.9144,
+      "step": 1141
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3343036645868432e-06,
+      "loss": 1.6482,
+      "step": 1142
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.331741329093483e-06,
+      "loss": 1.1715,
+      "step": 1143
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.329178010738678e-06,
+      "loss": 0.9454,
+      "step": 1144
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3266137125514784e-06,
+      "loss": 0.7856,
+      "step": 1145
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.324048437562094e-06,
+      "loss": 1.0394,
+      "step": 1146
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.321482188801888e-06,
+      "loss": 0.8741,
+      "step": 1147
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3189149693033717e-06,
+      "loss": 1.2134,
+      "step": 1148
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3163467821002082e-06,
+      "loss": 1.1549,
+      "step": 1149
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3137776302272017e-06,
+      "loss": 1.6155,
+      "step": 1150
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3112075167202955e-06,
+      "loss": 1.8034,
+      "step": 1151
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.3086364446165714e-06,
+      "loss": 0.9134,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.306064416954241e-06,
+      "loss": 0.8948,
+      "step": 1153
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.3034914367726483e-06,
+      "loss": 1.4639,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.300917507112261e-06,
+      "loss": 1.3632,
+      "step": 1155
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.2983426310146698e-06,
+      "loss": 1.0861,
+      "step": 1156
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.2957668115225828e-06,
+      "loss": 1.1023,
+      "step": 1157
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.2931900516798232e-06,
+      "loss": 1.3424,
+      "step": 1158
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.290612354531325e-06,
+      "loss": 1.9575,
+      "step": 1159
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.2880337231231312e-06,
+      "loss": 1.9992,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 1.2133097648620605,
+      "eval_runtime": 53.8686,
+      "eval_samples_per_second": 14.331,
+      "eval_steps_per_second": 7.166,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.2854541605023882e-06,
+      "loss": 1.2993,
+      "step": 1161
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.282873669717342e-06,
+      "loss": 1.3915,
+      "step": 1162
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.280292253817336e-06,
+      "loss": 1.1089,
+      "step": 1163
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.2777099158528064e-06,
+      "loss": 1.7346,
+      "step": 1164
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.27512665887528e-06,
+      "loss": 1.5882,
+      "step": 1165
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 1.3054,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.269957400092767e-06,
+      "loss": 1.148,
+      "step": 1167
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.267371404396249e-06,
+      "loss": 1.0103,
+      "step": 1168
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.2647845019036622e-06,
+      "loss": 1.2608,
+      "step": 1169
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.2621966956719265e-06,
+      "loss": 0.7083,
+      "step": 1170
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.2596079887590295e-06,
+      "loss": 1.0038,
+      "step": 1171
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.2570183842240255e-06,
+      "loss": 0.9609,
+      "step": 1172
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.254427885127025e-06,
+      "loss": 0.8564,
+      "step": 1173
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.2518364945292003e-06,
+      "loss": 1.177,
+      "step": 1174
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.249244215492774e-06,
+      "loss": 1.707,
+      "step": 1175
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.2466510510810185e-06,
+      "loss": 1.1875,
+      "step": 1176
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.2440570043582554e-06,
+      "loss": 1.245,
+      "step": 1177
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.241462078389846e-06,
+      "loss": 1.8574,
+      "step": 1178
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.2388662762421914e-06,
+      "loss": 1.4792,
+      "step": 1179
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.236269600982729e-06,
+      "loss": 1.2552,
+      "step": 1180
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.233672055679926e-06,
+      "loss": 1.1056,
+      "step": 1181
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2310736434032797e-06,
+      "loss": 1.524,
+      "step": 1182
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.228474367223312e-06,
+      "loss": 0.8447,
+      "step": 1183
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2258742302115632e-06,
+      "loss": 1.8112,
+      "step": 1184
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2232732354405936e-06,
+      "loss": 0.8166,
+      "step": 1185
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2206713859839755e-06,
+      "loss": 1.0981,
+      "step": 1186
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.218068684916291e-06,
+      "loss": 0.789,
+      "step": 1187
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2154651353131305e-06,
+      "loss": 1.8448,
+      "step": 1188
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.212860740251085e-06,
+      "loss": 1.0422,
+      "step": 1189
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.210255502807745e-06,
+      "loss": 1.169,
+      "step": 1190
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2076494260616982e-06,
+      "loss": 1.2136,
+      "step": 1191
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.20504251309252e-06,
+      "loss": 1.3436,
+      "step": 1192
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.202434766980779e-06,
+      "loss": 0.7606,
+      "step": 1193
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.199826190808026e-06,
+      "loss": 0.987,
+      "step": 1194
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.1972167876567907e-06,
+      "loss": 1.3677,
+      "step": 1195
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.1946065606105832e-06,
+      "loss": 1.3028,
+      "step": 1196
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.1919955127538863e-06,
+      "loss": 0.9541,
+      "step": 1197
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.189383647172152e-06,
+      "loss": 1.2335,
+      "step": 1198
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.1867709669517983e-06,
+      "loss": 1.3703,
+      "step": 1199
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.184157475180208e-06,
+      "loss": 1.3037,
+      "step": 1200
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.181543174945719e-06,
+      "loss": 1.2641,
+      "step": 1201
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.17892806933763e-06,
+      "loss": 1.0842,
+      "step": 1202
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.176312161446187e-06,
+      "loss": 0.8864,
+      "step": 1203
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.1736954543625843e-06,
+      "loss": 1.7271,
+      "step": 1204
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.1710779511789634e-06,
+      "loss": 1.0044,
+      "step": 1205
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.1684596549884047e-06,
+      "loss": 0.8295,
+      "step": 1206
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.1658405688849246e-06,
+      "loss": 0.737,
+      "step": 1207
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.163220695963476e-06,
+      "loss": 0.7618,
+      "step": 1208
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.1606000393199393e-06,
+      "loss": 1.5471,
+      "step": 1209
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.1579786020511203e-06,
+      "loss": 0.7496,
+      "step": 1210
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.15535638725475e-06,
+      "loss": 1.707,
+      "step": 1211
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1527333980294748e-06,
+      "loss": 1.1097,
+      "step": 1212
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.15010963747486e-06,
+      "loss": 1.6137,
+      "step": 1213
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1474851086913793e-06,
+      "loss": 1.0059,
+      "step": 1214
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1448598147804162e-06,
+      "loss": 1.6696,
+      "step": 1215
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1422337588442577e-06,
+      "loss": 0.8577,
+      "step": 1216
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1396069439860894e-06,
+      "loss": 0.7763,
+      "step": 1217
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.136979373309997e-06,
+      "loss": 1.5286,
+      "step": 1218
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.134351049920957e-06,
+      "loss": 1.3097,
+      "step": 1219
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1317219769248373e-06,
+      "loss": 1.4764,
+      "step": 1220
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.129092157428389e-06,
+      "loss": 1.2568,
+      "step": 1221
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1264615945392473e-06,
+      "loss": 1.2939,
+      "step": 1222
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1238302913659254e-06,
+      "loss": 1.1002,
+      "step": 1223
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1211982510178114e-06,
+      "loss": 1.2296,
+      "step": 1224
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1185654766051638e-06,
+      "loss": 0.9342,
+      "step": 1225
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.115931971239109e-06,
+      "loss": 1.8502,
+      "step": 1226
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.1132977380316388e-06,
+      "loss": 0.9097,
+      "step": 1227
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.1106627800956014e-06,
+      "loss": 1.0691,
+      "step": 1228
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.1080271005447038e-06,
+      "loss": 1.3972,
+      "step": 1229
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.105390702493508e-06,
+      "loss": 0.8666,
+      "step": 1230
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.1027535890574194e-06,
+      "loss": 1.8229,
+      "step": 1231
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.100115763352693e-06,
+      "loss": 0.7901,
+      "step": 1232
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.097477228496425e-06,
+      "loss": 1.1264,
+      "step": 1233
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.094837987606547e-06,
+      "loss": 1.2298,
+      "step": 1234
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0921980438018293e-06,
+      "loss": 1.4231,
+      "step": 1235
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.089557400201869e-06,
+      "loss": 1.0321,
+      "step": 1236
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0869160599270915e-06,
+      "loss": 0.8622,
+      "step": 1237
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0842740260987457e-06,
+      "loss": 1.2512,
+      "step": 1238
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0816313018389e-06,
+      "loss": 1.0031,
+      "step": 1239
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.0789878902704385e-06,
+      "loss": 0.9935,
+      "step": 1240
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.0763437945170576e-06,
+      "loss": 1.0102,
+      "step": 1241
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.0736990177032614e-06,
+      "loss": 1.83,
+      "step": 1242
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.0710535629543597e-06,
+      "loss": 1.5906,
+      "step": 1243
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.0684074333964642e-06,
+      "loss": 1.0378,
+      "step": 1244
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.065760632156482e-06,
+      "loss": 0.8305,
+      "step": 1245
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.063113162362115e-06,
+      "loss": 0.7797,
+      "step": 1246
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.0604650271418556e-06,
+      "loss": 1.111,
+      "step": 1247
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.0578162296249812e-06,
+      "loss": 0.7282,
+      "step": 1248
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.055166772941554e-06,
+      "loss": 1.1565,
+      "step": 1249
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.052516660222412e-06,
+      "loss": 1.269,
+      "step": 1250
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.049865894599172e-06,
+      "loss": 1.3521,
+      "step": 1251
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.0472144792042196e-06,
+      "loss": 1.4313,
+      "step": 1252
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.0445624171707085e-06,
+      "loss": 0.766,
+      "step": 1253
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.041909711632558e-06,
+      "loss": 0.7068,
+      "step": 1254
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0392563657244472e-06,
+      "loss": 0.7982,
+      "step": 1255
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0366023825818107e-06,
+      "loss": 0.925,
+      "step": 1256
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0339477653408382e-06,
+      "loss": 0.8127,
+      "step": 1257
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.031292517138468e-06,
+      "loss": 1.2297,
+      "step": 1258
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0286366411123817e-06,
+      "loss": 0.8651,
+      "step": 1259
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.025980140401007e-06,
+      "loss": 0.7399,
+      "step": 1260
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.023323018143506e-06,
+      "loss": 0.8387,
+      "step": 1261
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0206652774797775e-06,
+      "loss": 0.9132,
+      "step": 1262
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0180069215504505e-06,
+      "loss": 1.276,
+      "step": 1263
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0153479534968792e-06,
+      "loss": 1.1802,
+      "step": 1264
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0126883764611452e-06,
+      "loss": 1.2264,
+      "step": 1265
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0100281935860465e-06,
+      "loss": 1.0068,
+      "step": 1266
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0073674080150977e-06,
+      "loss": 1.4417,
+      "step": 1267
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.0047060228925256e-06,
+      "loss": 1.3083,
+      "step": 1268
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.002044041363267e-06,
+      "loss": 1.3119,
+      "step": 1269
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9993814665729605e-06,
+      "loss": 1.5967,
+      "step": 1270
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.996718301667948e-06,
+      "loss": 0.8924,
+      "step": 1271
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9940545497952687e-06,
+      "loss": 1.01,
+      "step": 1272
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.991390214102654e-06,
+      "loss": 1.007,
+      "step": 1273
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9887252977385272e-06,
+      "loss": 1.0263,
+      "step": 1274
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9860598038519954e-06,
+      "loss": 0.8872,
+      "step": 1275
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.98339373559285e-06,
+      "loss": 1.9292,
+      "step": 1276
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9807270961115603e-06,
+      "loss": 1.2176,
+      "step": 1277
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.978059888559271e-06,
+      "loss": 1.0038,
+      "step": 1278
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9753921160877976e-06,
+      "loss": 1.0877,
+      "step": 1279
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.972723781849624e-06,
+      "loss": 0.9189,
+      "step": 1280
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.970054888997897e-06,
+      "loss": 0.7527,
+      "step": 1281
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9673854406864244e-06,
+      "loss": 0.9119,
+      "step": 1282
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.964715440069669e-06,
+      "loss": 0.7733,
+      "step": 1283
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9620448903027484e-06,
+      "loss": 0.7463,
+      "step": 1284
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9593737945414264e-06,
+      "loss": 0.6778,
+      "step": 1285
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9567021559421144e-06,
+      "loss": 1.4462,
+      "step": 1286
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9540299776618637e-06,
+      "loss": 1.5849,
+      "step": 1287
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9513572628583654e-06,
+      "loss": 1.5328,
+      "step": 1288
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9486840146899414e-06,
+      "loss": 0.8186,
+      "step": 1289
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.946010236315546e-06,
+      "loss": 1.144,
+      "step": 1290
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.943335930894762e-06,
+      "loss": 0.7505,
+      "step": 1291
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.940661101587789e-06,
+      "loss": 1.2277,
+      "step": 1292
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.937985751555452e-06,
+      "loss": 1.2852,
+      "step": 1293
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.935309883959188e-06,
+      "loss": 1.6152,
+      "step": 1294
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9326335019610465e-06,
+      "loss": 0.7664,
+      "step": 1295
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9299566087236834e-06,
+      "loss": 1.2087,
+      "step": 1296
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9272792074103616e-06,
+      "loss": 1.4253,
+      "step": 1297
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9246013011849415e-06,
+      "loss": 0.9122,
+      "step": 1298
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.921922893211883e-06,
+      "loss": 0.9546,
+      "step": 1299
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9192439866562363e-06,
+      "loss": 1.6417,
+      "step": 1300
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.916564584683641e-06,
+      "loss": 1.1093,
+      "step": 1301
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.913884690460325e-06,
+      "loss": 1.406,
+      "step": 1302
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9112043071530942e-06,
+      "loss": 1.6273,
+      "step": 1303
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9085234379293353e-06,
+      "loss": 1.4468,
+      "step": 1304
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.905842085957007e-06,
+      "loss": 0.8916,
+      "step": 1305
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 1.2053191661834717,
+      "eval_runtime": 53.6929,
+      "eval_samples_per_second": 14.378,
+      "eval_steps_per_second": 7.189,
+      "step": 1305
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9031602544046395e-06,
+      "loss": 1.1239,
+      "step": 1306
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9004779464413303e-06,
+      "loss": 0.9879,
+      "step": 1307
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.8977951652367382e-06,
+      "loss": 0.7417,
+      "step": 1308
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.895111913961083e-06,
+      "loss": 1.4614,
+      "step": 1309
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.8924281957851398e-06,
+      "loss": 0.8142,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.8897440138802336e-06,
+      "loss": 0.9211,
+      "step": 1311
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.8870593714182394e-06,
+      "loss": 1.1745,
+      "step": 1312
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8843742715715755e-06,
+      "loss": 0.813,
+      "step": 1313
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8816887175132015e-06,
+      "loss": 1.3786,
+      "step": 1314
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8790027124166114e-06,
+      "loss": 1.2715,
+      "step": 1315
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8763162594558364e-06,
+      "loss": 1.0655,
+      "step": 1316
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8736293618054318e-06,
+      "loss": 1.4229,
+      "step": 1317
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8709420226404828e-06,
+      "loss": 0.8199,
+      "step": 1318
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8682542451365943e-06,
+      "loss": 0.7968,
+      "step": 1319
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8655660324698887e-06,
+      "loss": 2.126,
+      "step": 1320
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8628773878170048e-06,
+      "loss": 1.0297,
+      "step": 1321
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.860188314355088e-06,
+      "loss": 1.6878,
+      "step": 1322
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.857498815261795e-06,
+      "loss": 0.9881,
+      "step": 1323
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.854808893715283e-06,
+      "loss": 0.9988,
+      "step": 1324
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8521185528942086e-06,
+      "loss": 1.1359,
+      "step": 1325
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.8494277959777233e-06,
+      "loss": 0.9327,
+      "step": 1326
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8467366261454723e-06,
+      "loss": 0.8055,
+      "step": 1327
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8440450465775853e-06,
+      "loss": 1.164,
+      "step": 1328
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.84135306045468e-06,
+      "loss": 0.7184,
+      "step": 1329
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8386606709578535e-06,
+      "loss": 1.9446,
+      "step": 1330
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.835967881268677e-06,
+      "loss": 1.1398,
+      "step": 1331
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8332746945691976e-06,
+      "loss": 1.3205,
+      "step": 1332
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8305811140419298e-06,
+      "loss": 0.8334,
+      "step": 1333
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8278871428698544e-06,
+      "loss": 0.9549,
+      "step": 1334
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8251927842364147e-06,
+      "loss": 0.8567,
+      "step": 1335
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.822498041325509e-06,
+      "loss": 0.8521,
+      "step": 1336
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8198029173214913e-06,
+      "loss": 0.9907,
+      "step": 1337
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8171074154091677e-06,
+      "loss": 1.3401,
+      "step": 1338
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.814411538773788e-06,
+      "loss": 1.7585,
+      "step": 1339
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.811715290601046e-06,
+      "loss": 0.8718,
+      "step": 1340
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.8090186740770752e-06,
+      "loss": 1.3148,
+      "step": 1341
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.806321692388443e-06,
+      "loss": 1.5861,
+      "step": 1342
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.8036243487221506e-06,
+      "loss": 0.6952,
+      "step": 1343
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.800926646265623e-06,
+      "loss": 0.6928,
+      "step": 1344
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.798228588206712e-06,
+      "loss": 1.435,
+      "step": 1345
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7955301777336926e-06,
+      "loss": 1.2767,
+      "step": 1346
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.792831418035248e-06,
+      "loss": 0.8592,
+      "step": 1347
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7901323123004816e-06,
+      "loss": 1.371,
+      "step": 1348
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7874328637189028e-06,
+      "loss": 0.7175,
+      "step": 1349
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7847330754804257e-06,
+      "loss": 1.0062,
+      "step": 1350
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.782032950775367e-06,
+      "loss": 0.6584,
+      "step": 1351
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7793324927944398e-06,
+      "loss": 1.0581,
+      "step": 1352
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.776631704728752e-06,
+      "loss": 1.4408,
+      "step": 1353
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7739305897698023e-06,
+      "loss": 1.4966,
+      "step": 1354
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7712291511094725e-06,
+      "loss": 0.9441,
+      "step": 1355
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7685273919400306e-06,
+      "loss": 0.8899,
+      "step": 1356
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.765825315454121e-06,
+      "loss": 1.4834,
+      "step": 1357
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7631229248447637e-06,
+      "loss": 1.6439,
+      "step": 1358
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.76042022330535e-06,
+      "loss": 1.8315,
+      "step": 1359
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.75771721402964e-06,
+      "loss": 1.636,
+      "step": 1360
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7550139002117543e-06,
+      "loss": 1.1515,
+      "step": 1361
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.752310285046176e-06,
+      "loss": 1.0332,
+      "step": 1362
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.749606371727744e-06,
+      "loss": 2.0102,
+      "step": 1363
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7469021634516475e-06,
+      "loss": 1.3126,
+      "step": 1364
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7441976634134275e-06,
+      "loss": 0.9242,
+      "step": 1365
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.741492874808967e-06,
+      "loss": 1.3781,
+      "step": 1366
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7387878008344902e-06,
+      "loss": 0.7837,
+      "step": 1367
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7360824446865615e-06,
+      "loss": 1.571,
+      "step": 1368
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7333768095620745e-06,
+      "loss": 1.6734,
+      "step": 1369
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.730670898658255e-06,
+      "loss": 1.0149,
+      "step": 1370
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.727964715172656e-06,
+      "loss": 1.0424,
+      "step": 1371
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.725258262303148e-06,
+      "loss": 1.8595,
+      "step": 1372
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7225515432479245e-06,
+      "loss": 0.9829,
+      "step": 1373
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7198445612054914e-06,
+      "loss": 1.548,
+      "step": 1374
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7171373193746653e-06,
+      "loss": 0.9945,
+      "step": 1375
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.714429820954571e-06,
+      "loss": 0.6336,
+      "step": 1376
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.711722069144635e-06,
+      "loss": 1.2451,
+      "step": 1377
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7090140671445836e-06,
+      "loss": 1.7466,
+      "step": 1378
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7063058181544395e-06,
+      "loss": 1.4181,
+      "step": 1379
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7035973253745167e-06,
+      "loss": 1.3393,
+      "step": 1380
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.700888592005417e-06,
+      "loss": 1.2597,
+      "step": 1381
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.698179621248028e-06,
+      "loss": 1.5118,
+      "step": 1382
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.695470416303516e-06,
+      "loss": 0.9365,
+      "step": 1383
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6927609803733245e-06,
+      "loss": 0.9296,
+      "step": 1384
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6900513166591715e-06,
+      "loss": 0.9604,
+      "step": 1385
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6873414283630405e-06,
+      "loss": 0.9642,
+      "step": 1386
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.684631318687185e-06,
+      "loss": 0.8974,
+      "step": 1387
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6819209908341176e-06,
+      "loss": 1.0633,
+      "step": 1388
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.679210448006607e-06,
+      "loss": 1.8002,
+      "step": 1389
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6764996934076805e-06,
+      "loss": 1.3501,
+      "step": 1390
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6737887302406117e-06,
+      "loss": 0.75,
+      "step": 1391
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.671077561708922e-06,
+      "loss": 0.7531,
+      "step": 1392
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.668366191016376e-06,
+      "loss": 0.7146,
+      "step": 1393
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6656546213669764e-06,
+      "loss": 0.8559,
+      "step": 1394
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.662942855964961e-06,
+      "loss": 1.399,
+      "step": 1395
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6602308980148007e-06,
+      "loss": 1.1126,
+      "step": 1396
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.65751875072119e-06,
+      "loss": 1.6005,
+      "step": 1397
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.6548064172890524e-06,
+      "loss": 1.2356,
+      "step": 1398
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.652093900923527e-06,
+      "loss": 1.4258,
+      "step": 1399
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.64938120482997e-06,
+      "loss": 1.2329,
+      "step": 1400
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6466683322139515e-06,
+      "loss": 1.639,
+      "step": 1401
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6439552862812495e-06,
+      "loss": 1.2395,
+      "step": 1402
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.641242070237846e-06,
+      "loss": 1.0329,
+      "step": 1403
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.638528687289925e-06,
+      "loss": 1.3817,
+      "step": 1404
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6358151406438675e-06,
+      "loss": 1.3485,
+      "step": 1405
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6331014335062477e-06,
+      "loss": 0.7669,
+      "step": 1406
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.63038756908383e-06,
+      "loss": 0.7133,
+      "step": 1407
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6276735505835636e-06,
+      "loss": 1.159,
+      "step": 1408
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.62495938121258e-06,
+      "loss": 0.7357,
+      "step": 1409
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6222450641781904e-06,
+      "loss": 0.7507,
+      "step": 1410
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6195306026878788e-06,
+      "loss": 1.1725,
+      "step": 1411
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6168159999492997e-06,
+      "loss": 1.224,
+      "step": 1412
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.614101259170278e-06,
+      "loss": 0.8986,
+      "step": 1413
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6113863835587965e-06,
+      "loss": 1.2494,
+      "step": 1414
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.608671376323002e-06,
+      "loss": 1.0252,
+      "step": 1415
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.605956240671193e-06,
+      "loss": 0.9566,
+      "step": 1416
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.6032409798118226e-06,
+      "loss": 0.9629,
+      "step": 1417
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.60052559695349e-06,
+      "loss": 0.6855,
+      "step": 1418
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5978100953049406e-06,
+      "loss": 1.2674,
+      "step": 1419
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5950944780750577e-06,
+      "loss": 0.8772,
+      "step": 1420
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.592378748472863e-06,
+      "loss": 0.8036,
+      "step": 1421
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.58966290970751e-06,
+      "loss": 1.1676,
+      "step": 1422
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.586946964988281e-06,
+      "loss": 1.8266,
+      "step": 1423
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.584230917524586e-06,
+      "loss": 1.6392,
+      "step": 1424
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.581514770525953e-06,
+      "loss": 0.8578,
+      "step": 1425
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.57879852720203e-06,
+      "loss": 0.7239,
+      "step": 1426
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5760821907625767e-06,
+      "loss": 1.2385,
+      "step": 1427
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.573365764417465e-06,
+      "loss": 1.4696,
+      "step": 1428
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.570649251376673e-06,
+      "loss": 1.7116,
+      "step": 1429
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.567932654850278e-06,
+      "loss": 1.3335,
+      "step": 1430
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5652159780484598e-06,
+      "loss": 0.9493,
+      "step": 1431
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5624992241814922e-06,
+      "loss": 1.4103,
+      "step": 1432
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.559782396459738e-06,
+      "loss": 1.6927,
+      "step": 1433
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5570654980936493e-06,
+      "loss": 1.5447,
+      "step": 1434
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5543485322937617e-06,
+      "loss": 0.7246,
+      "step": 1435
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5516315022706893e-06,
+      "loss": 0.86,
+      "step": 1436
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5489144112351234e-06,
+      "loss": 1.8581,
+      "step": 1437
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.546197262397825e-06,
+      "loss": 0.8806,
+      "step": 1438
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5434800589696263e-06,
+      "loss": 0.8996,
+      "step": 1439
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.5407628041614234e-06,
+      "loss": 0.9795,
+      "step": 1440
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.538045501184172e-06,
+      "loss": 0.8858,
+      "step": 1441
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.535328153248884e-06,
+      "loss": 1.1132,
+      "step": 1442
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.532610763566628e-06,
+      "loss": 1.5197,
+      "step": 1443
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5298933353485174e-06,
+      "loss": 1.2823,
+      "step": 1444
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5271758718057138e-06,
+      "loss": 2.0808,
+      "step": 1445
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.524458376149421e-06,
+      "loss": 1.4084,
+      "step": 1446
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.521740851590879e-06,
+      "loss": 1.7921,
+      "step": 1447
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.519023301341364e-06,
+      "loss": 1.6181,
+      "step": 1448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.516305728612179e-06,
+      "loss": 1.2924,
+      "step": 1449
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5135881366146585e-06,
+      "loss": 0.7885,
+      "step": 1450
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.1991530656814575,
+      "eval_runtime": 53.8047,
+      "eval_samples_per_second": 14.348,
+      "eval_steps_per_second": 7.174,
+      "step": 1450
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5108705285601554e-06,
+      "loss": 1.5017,
+      "step": 1451
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.508152907660044e-06,
+      "loss": 0.7704,
+      "step": 1452
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.505435277125713e-06,
+      "loss": 1.4009,
+      "step": 1453
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5027176401685628e-06,
+      "loss": 1.5988,
+      "step": 1454
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.5e-06,
+      "loss": 1.0798,
+      "step": 1455
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.497282359831438e-06,
+      "loss": 1.5089,
+      "step": 1456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4945647228742883e-06,
+      "loss": 1.5705,
+      "step": 1457
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.491847092339957e-06,
+      "loss": 1.012,
+      "step": 1458
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4891294714398455e-06,
+      "loss": 0.7862,
+      "step": 1459
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4864118633853424e-06,
+      "loss": 1.3208,
+      "step": 1460
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.483694271387821e-06,
+      "loss": 1.3518,
+      "step": 1461
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.480976698658637e-06,
+      "loss": 1.2256,
+      "step": 1462
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.478259148409121e-06,
+      "loss": 1.5057,
+      "step": 1463
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4755416238505792e-06,
+      "loss": 0.6883,
+      "step": 1464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.472824128194286e-06,
+      "loss": 1.4382,
+      "step": 1465
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4701066646514843e-06,
+      "loss": 1.4933,
+      "step": 1466
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4673892364333736e-06,
+      "loss": 1.3664,
+      "step": 1467
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4646718467511167e-06,
+      "loss": 1.7627,
+      "step": 1468
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4619544988158293e-06,
+      "loss": 1.3693,
+      "step": 1469
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.459237195838577e-06,
+      "loss": 0.9821,
+      "step": 1470
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.456519941030374e-06,
+      "loss": 1.3148,
+      "step": 1471
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.453802737602176e-06,
+      "loss": 1.0727,
+      "step": 1472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4510855887648775e-06,
+      "loss": 0.9669,
+      "step": 1473
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4483684977293116e-06,
+      "loss": 1.3828,
+      "step": 1474
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.445651467706239e-06,
+      "loss": 1.3555,
+      "step": 1475
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4429345019063516e-06,
+      "loss": 0.8507,
+      "step": 1476
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.440217603540263e-06,
+      "loss": 0.6893,
+      "step": 1477
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4375007758185086e-06,
+      "loss": 1.532,
+      "step": 1478
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.434784021951541e-06,
+      "loss": 0.726,
+      "step": 1479
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.432067345149723e-06,
+      "loss": 0.5863,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.429350748623328e-06,
+      "loss": 1.0244,
+      "step": 1481
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4266342355825354e-06,
+      "loss": 0.719,
+      "step": 1482
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4239178092374237e-06,
+      "loss": 0.7407,
+      "step": 1483
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.421201472797971e-06,
+      "loss": 1.3884,
+      "step": 1484
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4184852294740474e-06,
+      "loss": 0.9381,
+      "step": 1485
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.415769082475414e-06,
+      "loss": 0.6054,
+      "step": 1486
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4130530350117194e-06,
+      "loss": 1.0417,
+      "step": 1487
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4103370902924916e-06,
+      "loss": 0.645,
+      "step": 1488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4076212515271384e-06,
+      "loss": 0.4432,
+      "step": 1489
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4049055219249435e-06,
+      "loss": 0.9322,
+      "step": 1490
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4021899046950607e-06,
+      "loss": 0.5862,
+      "step": 1491
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3994744030465106e-06,
+      "loss": 1.3218,
+      "step": 1492
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3967590201881786e-06,
+      "loss": 0.5756,
+      "step": 1493
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3940437593288075e-06,
+      "loss": 1.3496,
+      "step": 1494
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3913286236769988e-06,
+      "loss": 0.8932,
+      "step": 1495
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.388613616441204e-06,
+      "loss": 0.8556,
+      "step": 1496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3858987408297225e-06,
+      "loss": 1.4038,
+      "step": 1497
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3831840000507007e-06,
+      "loss": 1.1387,
+      "step": 1498
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3804693973121225e-06,
+      "loss": 0.6889,
+      "step": 1499
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3777549358218105e-06,
+      "loss": 0.6566,
+      "step": 1500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2900,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "total_flos": 5.24254747557888e+17,
+  "trial_name": null,
+  "trial_params": null
+}