{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.850632911392405,
  "eval_steps": 500,
  "global_step": 224,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.030379746835443037,
      "grad_norm": 6.006213471081438,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.7917,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.2254607837063265,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.8352,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.033293241527638,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.8056,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 5.6177843051297724,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.7998,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.723227334218072,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.7966,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.5374960923382606,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.7432,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.126329485485419,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.738,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 3.873369258557868,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.7192,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 3.899288948622671,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.698,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 4.121150199216906,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.6933,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 4.253897850285379,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.6662,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 3.503928981732218,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.641,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 2.1797633391983697,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.6131,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.9144365807987356,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.6092,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 2.4090527409072395,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.5987,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 2.1201232027340455,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.561,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.2244003008036553,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.5766,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.146399280795341,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.566,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 1.3934479350243683,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.5672,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.974415129473555,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.599,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 0.9919902220679716,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.529,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 1.040576121443834,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.5438,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 0.7352410276728014,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 0.828148529201473,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.5525,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.8269176543378692,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.5144,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.6869010086082622,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.5047,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.7689050169156533,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.5478,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.6354264694082578,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.5212,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.6922907069132699,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.5033,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.6323752752121719,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.5258,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.5593183327477625,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.4979,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.6622551856194377,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.5248,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 0.6764945387024673,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.6827,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.6561499543200164,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.469,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.5298141480802675,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.4372,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.5505361666397994,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.4661,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.5105637977979922,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.3946,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.6434975406027199,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.5268,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.6008580032405805,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.4601,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.5959585981147149,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.4591,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.546478161597411,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.4329,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.6283874355809709,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.4572,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.5615002194011511,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.4369,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.6258423268174595,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.4316,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.5534904501954303,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.4249,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.669119095568495,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.4741,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.5093735104578105,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.4113,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.6286618048992411,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.4209,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.686238431846376,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.4852,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.5083899652707847,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.3972,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.5598850359421916,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.4913,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.5766825647521311,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.4621,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.5086207648877343,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.418,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.5873176022167675,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.4546,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.5303690045849334,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.4214,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.49337675464597536,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.4144,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.5223857062364483,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.4226,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.4568296082068541,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.3996,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.548700809584534,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.4603,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.4657778954869675,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.4323,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.5585461501887276,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.4237,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.5017239451917875,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.3994,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.48711586534451035,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.4519,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.48646155441739797,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.4467,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.6897580966520201,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.5046,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.5631698052635786,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.4692,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.6468306572972523,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.3942,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 0.5218203561260673,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.3286,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.7071640994190285,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.3418,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.5968775859817312,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.3328,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.6417220025244813,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.3394,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.5311259420709453,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.3315,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.6830204026929048,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.3865,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.47877101908626185,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.3154,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.4643666155507386,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.3249,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.5212861098993654,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.3314,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.5422090304154028,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.3774,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.40218697770413736,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.2784,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.6589258015531194,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.383,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.5430798030074878,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.3565,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.5073906623339022,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.3569,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.5050027438908327,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.3209,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.4765935907526067,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.3297,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.46207789314346887,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.3044,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.5752561163345511,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.4127,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.5137987799240195,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.3065,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.49228750648377373,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.3341,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.6814522876949816,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.3202,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.537936998473839,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.3673,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.4995308807561755,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.3352,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.5042144418218611,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.3649,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.4810205529248607,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.3465,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.46777668763063673,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.3312,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.46595629768395286,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.3389,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.40013868400765945,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.2892,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.524326098641433,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.3889,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.4417005423165919,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.3382,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.6032209387187255,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.4444,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.6573238334500671,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.2592,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.5605144807167273,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.2547,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.6531233040095235,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.273,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.9334210522962539,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.2575,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.5559069460326249,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.2552,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.6492722466526953,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.2552,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.5510778935673912,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.2392,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.5525434722803316,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.2267,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.5405856149923781,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.2518,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.5645935517029098,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.2554,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.4962295976875595,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.2353,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.5724199030679713,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.244,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.5112070197263711,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.2399,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.5440312743755955,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.2323,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.48627544442489706,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.2554,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.5021444870802758,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.2401,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.46169404742804154,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.2656,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.48245585451376644,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.2516,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.4598449307989322,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.2543,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.4479409796343516,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.2456,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.45462564739248734,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.257,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.4015595416578657,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.2309,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.45494954606499216,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.2567,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.41059230945366637,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.2461,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.46432852862870677,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.266,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.40805606299951863,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.2462,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.43921004324299806,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.2473,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.44473479159040624,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.2449,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.44748351943942116,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.2557,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.518560853267675,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.2663,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.4246648781720374,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.2361,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.41423647843411554,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.2569,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.6163458043061125,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.3476,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.5914798912668137,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.1681,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.44321978876286244,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.1743,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.4646965237711949,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.174,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.7248155117716496,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.1844,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.6109335002816132,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.2071,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.42282157241586654,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.1633,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.5014368521052023,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.1691,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.4837752367217211,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.184,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.43213740691333613,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.1629,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.403652454589129,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.1738,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.421582562547759,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.1742,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.44518835218880165,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.1832,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.4546106780236523,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.1682,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.43761106153770857,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.164,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.4310227803893202,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.168,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.45481355409412405,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.174,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.3957283319952441,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.1665,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.4552678978797154,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.2019,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.3928335896952226,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.1534,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.4015668482093742,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.1527,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.4204491043436094,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.1818,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.3731008292524761,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.1666,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.4133162907889197,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.1803,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.43836732632578446,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.1732,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.4191158440787698,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.178,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.3761936549445521,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.1566,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.3741682566256842,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.1735,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.40744096092794163,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.1789,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.3732566845318827,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.1603,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.4104311075804241,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.1741,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.4174241773674738,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.1932,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.4960747398972354,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.2334,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.49311557726904276,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.1724,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.4219724136306463,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.1182,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.3575965523976325,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.1102,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.3819441296997781,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.1421,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.36755760720758723,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.1144,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.4970308078631468,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.1236,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.5309873101289584,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.1456,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.4603253927514517,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.1286,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.3952284019691279,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.1355,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.362838535477914,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.1131,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.3945869010422975,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.1419,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.4052553677032924,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.1366,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.3457865463376092,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.1095,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.40994146296011247,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.1325,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.35229875639507074,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.1153,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.3501790972555334,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.131,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.33458569532699606,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.1127,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.3613754567991387,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.1315,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.36482185521540356,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.1264,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.3537624081643742,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.1058,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.41078850995506017,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.1349,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.3415708757919234,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.0931,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.3738318856202389,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.1324,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.3201115593014112,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.1155,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.3193124066196814,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.1266,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.352658463840024,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.1401,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.35577963695957643,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.1179,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.3046690367490559,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.1041,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.3253967100229279,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.1193,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.3355585754343889,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.1186,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.347764998604189,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.1214,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.3623556898310986,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.1428,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.4545442024513848,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.1609,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.387114634683258,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.0996,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.36921391920746444,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.1127,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.3642487093341055,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.1123,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.30930492840971036,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.106,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.32464631965147894,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.0966,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.29283184192001643,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.1012,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.3094164314460087,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.1017,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.3302230130000344,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.1053,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.3046311548899495,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.1128,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.31482728266911736,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.0992,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.3119001336132193,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.1047,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.3566508941703967,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.1037,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.3178470550954623,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.1025,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.3083171094133544,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.1022,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.30792921087733777,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.091,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.31286644898467275,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.0965,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.32575239630053976,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.1106,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.3164925601374238,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.1049,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.3006576850076976,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.1036,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.30959763059509177,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.1022,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.32718766109021413,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.1095,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.31611073824806707,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.0987,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.3226117308334687,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.1153,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.3188247426029841,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.1203,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.2918096040036267,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.101,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.3006475270572043,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.1063,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.3084749992680357,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.1022,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.30755613849148156,
      "learning_rate": 0.0,
      "loss": 0.108,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 3.3605374365505946e+17,
      "train_loss": 0.30083318214331356,
      "train_runtime": 14587.34,
      "train_samples_per_second": 1.516,
      "train_steps_per_second": 0.015
    }
  ],
  "logging_steps": 1,
  "max_steps": 224,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.3605374365505946e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}