{
  "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": 6.328099678449525,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.8202,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.297358235424027,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.8194,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.301558407733257,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.8266,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 5.711902603765234,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.7844,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 4.221837713593794,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.7604,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.259916793628621,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.7094,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 1.9150896667185284,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.6995,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 3.7504302335448942,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.7128,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 4.299369475549614,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.721,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 4.100245454975869,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.6968,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 4.045243716508803,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.679,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 3.30592773367142,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.6756,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 1.9028163551245074,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.6612,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 1.822367415324588,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.6345,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 1.929175393433327,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.625,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 1.447052334588171,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.5997,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 1.0104587873015645,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.5911,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 1.1811356996261588,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.6016,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 0.9302347784786774,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.5412,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 1.1063236632026656,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.5853,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 0.8391283894107245,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.5627,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 0.8081216436361469,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.5571,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 0.9104921955880105,
      "learning_rate": 2e-05,
      "loss": 0.5503,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 0.6782008168973742,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.5476,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.7074197142265733,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.5325,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 0.7053339414581664,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.5458,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 0.7106564667936467,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.5391,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 0.5394503329104179,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.528,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 0.5633367230219238,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.5081,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.6843923627902365,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.5208,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.4806483878725231,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.4985,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.5159623175987978,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.5106,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 0.8156979728940722,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.664,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.7208088057850367,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.5116,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.47902882807330865,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.4736,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.5468280769600697,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.4876,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.4870744083565367,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.4908,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.4980220974019647,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.4734,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.5609600583081555,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.484,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.4386050137552842,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.4565,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.5382082361564214,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.4947,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.448718159561216,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.4632,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 0.4680268435035972,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.4724,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.43578532675962506,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.4624,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.47014058991135543,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.4748,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.4286872900326167,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.4763,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.4249508421029805,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.4619,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.5426635438478858,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.4764,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.5715382321954325,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.4844,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.556353886166731,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.4798,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.40602427429233096,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.445,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.410992622154037,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.4595,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.46818703145861684,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.4489,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.3320524635048984,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.4669,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.3695647649201708,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.459,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.44458800129303727,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.4788,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.36669890958475326,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.4579,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.40317470267469624,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.4516,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.3830033673710046,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.4427,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.3610408811265048,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.4429,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.3981251279743908,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.4612,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.3486106354175202,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.4397,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.45709902609747943,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.472,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.3907671064553105,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.4524,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.4659390466720739,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.5218,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.4757923405813587,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.5351,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.4458163033921169,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.402,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 0.3961967002920918,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.4084,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.4166866686329854,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.4051,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.444043581663471,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.4056,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.4418822584821024,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.4182,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.513762793818981,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.3992,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.4524641435389744,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.4254,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.5261204916959731,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.4027,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.4540503907601427,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.4064,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.3529535177933412,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.4035,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.5166677816499226,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.4122,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.5278136735610292,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.4187,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.44253271962157364,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.3891,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.47732788637945245,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.4353,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.4871409614771859,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.3907,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.4607382186054202,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.4076,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.350513877265884,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.4039,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.45165951441755614,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.4046,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.42934074385619675,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.4109,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.341960993633312,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.3842,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.37186431805615333,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.3933,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.4333384904262265,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.4048,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.32394607365595307,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.3886,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 0.4010185098609762,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4106,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.4068330259382851,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.3912,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.37937109640783095,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.4163,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.49311740387664504,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.4087,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.3505395340432098,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.4061,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.38091464552263454,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.4101,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.3857714958847975,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.4063,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.2950043891018693,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.3929,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.4651445913505925,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.5141,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.5691872230966744,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.3662,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.39952247800464336,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.3585,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 0.3826703714414496,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.3586,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.5129651994957621,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.3688,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.4222951613050753,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.3559,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.3573808795754681,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.3539,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 0.48475589504514016,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.3569,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.384463803717113,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.3555,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.3943510843657851,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.3614,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.3953288926838918,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.3441,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.36861061059909456,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.3515,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.36700734058055634,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.3549,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.35432450602466337,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.3522,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.40098000170515286,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.339,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.371773613609398,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.3576,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.3812618794271841,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.3559,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.4211998306369166,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.3555,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.3237370226708903,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.3634,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.36303684110017476,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.3491,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.279372093963438,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.3373,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.3496593343858289,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.3563,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.2904478638039092,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.3508,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.31393096015037575,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.3532,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.32971110908004625,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.3546,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.2815667210970265,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.3654,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.324085242969168,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.357,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.3180405448754479,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.3577,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.3483493579672964,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.3479,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.2954658789786541,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.3601,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.3044202591486595,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.3472,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.3027909475475394,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.3478,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.31681040913324215,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.3578,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.3880197971513751,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.4492,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.3500089893201118,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.3131,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.31586925209306765,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.3283,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.37252514615104937,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.3165,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.3535284412890782,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.2995,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.31521012984486185,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.305,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.3192740990029793,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.3161,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.3303921884306372,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.3179,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.29689395281464653,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.3125,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.2981961569174302,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.3195,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.28896452694819574,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.3283,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.2955954566055393,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.3034,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.2616209389046269,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.313,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.2910593043854462,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.3146,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.2693015009531607,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.317,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.29708311735985876,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.3159,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.263556752957324,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.3017,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.264565783819196,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.3106,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.27189314678445364,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.3125,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.28331282396677754,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.3103,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.27717165214770556,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.3083,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.25104510674182184,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.3033,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.2772232641819828,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.3177,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.2513706037810617,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.3171,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.2375693711176268,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.3128,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.2784989605214622,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.3182,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.2629556556707916,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.3029,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.24406516753664173,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.3097,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.24241432056137993,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.3227,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.24911588817914426,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.2991,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.23963241338951638,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.3093,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.238350767239963,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.2984,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.2873358999434881,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.3563,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.33022561272849277,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.3274,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.3226928285291485,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.2904,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.28135674250563464,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.2833,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.26589471730576864,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.2785,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.3397931428453612,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.2707,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.34443599895848076,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.279,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.28502973603385,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.2805,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.2745535442161418,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.2926,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.2823596800843761,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.26,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.28998273548911163,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.3002,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.25736648266774537,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.2755,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.257137264361948,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.2739,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.2933555604190686,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.2976,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.27105775711801927,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.2676,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.2714298778465501,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.2934,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.2671847362096817,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.2823,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.24689291030673868,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.2734,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.2637213556975684,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.289,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.25020047597850015,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.2707,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.23779137054880234,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.2826,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.24093131103412868,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.2863,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.23807222189016014,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.2807,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.23755567816126796,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.2807,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.2428605572002419,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.2963,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.23361526713869593,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.2776,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.2323308877722213,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.2767,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.2362106108395449,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.2935,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.23801915518257494,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.2842,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.2186833299400843,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.2821,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.220396242182961,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.2856,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.21201066119863016,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.2794,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.22671925218629072,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.2951,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.2961777115923855,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.3675,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.26012035700928227,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.2795,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.24704520156881118,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.2679,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.23297706573524965,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.2693,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.23628242384199713,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.2735,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.21687242595159567,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.2705,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.21231922255439822,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.2663,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.21600842647150376,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.266,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.5342257013162511,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.2673,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.2231721836254536,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.2715,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.229133422018767,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.2669,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.21615234544627762,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.2601,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.23884142414590465,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.2711,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.22888957126530102,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.2679,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.22736752518229425,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.2856,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.20354841467535287,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.2532,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.20889091787617034,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.2593,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.21434684291534642,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.2766,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.21676023975071157,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.2725,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.21475211586189413,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.2698,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.22488374056232338,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.2635,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.21774035067897493,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.2762,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.21189327310036013,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.2695,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.20656641641610754,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.2623,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.21229553040623922,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.2772,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.20604379507259482,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.2652,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.20950661870734316,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.2688,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.21169061912460116,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.272,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.20989381506511842,
      "learning_rate": 0.0,
      "loss": 0.2637,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 8.799570510930248e+17,
      "train_loss": 0.39404168591967653,
      "train_runtime": 25889.0287,
      "train_samples_per_second": 0.854,
      "train_steps_per_second": 0.009
    }
  ],
  "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": 8.799570510930248e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}