{
  "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": 6.375513429053421,
      "learning_rate": 8.695652173913044e-07,
      "loss": 1.0529,
      "step": 1
    },
    {
      "epoch": 0.060759493670886074,
      "grad_norm": 6.646728126913626,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 1.062,
      "step": 2
    },
    {
      "epoch": 0.09113924050632911,
      "grad_norm": 6.74213427523718,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 1.0834,
      "step": 3
    },
    {
      "epoch": 0.12151898734177215,
      "grad_norm": 6.5239092218681085,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 1.0798,
      "step": 4
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 5.223436961421135,
      "learning_rate": 4.347826086956522e-06,
      "loss": 1.0477,
      "step": 5
    },
    {
      "epoch": 0.18227848101265823,
      "grad_norm": 2.466028784995944,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.9383,
      "step": 6
    },
    {
      "epoch": 0.21265822784810126,
      "grad_norm": 2.16088811137384,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.9567,
      "step": 7
    },
    {
      "epoch": 0.2430379746835443,
      "grad_norm": 4.577680515336394,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.9735,
      "step": 8
    },
    {
      "epoch": 0.27341772151898736,
      "grad_norm": 4.72999534804951,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.951,
      "step": 9
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 4.604185356938277,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.9202,
      "step": 10
    },
    {
      "epoch": 0.3341772151898734,
      "grad_norm": 4.4419944531614295,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.9406,
      "step": 11
    },
    {
      "epoch": 0.36455696202531646,
      "grad_norm": 4.131042413032702,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.8951,
      "step": 12
    },
    {
      "epoch": 0.3949367088607595,
      "grad_norm": 3.33579290893939,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.8762,
      "step": 13
    },
    {
      "epoch": 0.4253164556962025,
      "grad_norm": 3.031465534758255,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.8624,
      "step": 14
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 2.999946633497148,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.8344,
      "step": 15
    },
    {
      "epoch": 0.4860759493670886,
      "grad_norm": 2.7661320034701506,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.842,
      "step": 16
    },
    {
      "epoch": 0.5164556962025316,
      "grad_norm": 2.489006402151423,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.8138,
      "step": 17
    },
    {
      "epoch": 0.5468354430379747,
      "grad_norm": 2.3059289387035675,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.8267,
      "step": 18
    },
    {
      "epoch": 0.5772151898734177,
      "grad_norm": 1.7836216622884047,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.7877,
      "step": 19
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 1.9070648915077937,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.7527,
      "step": 20
    },
    {
      "epoch": 0.6379746835443038,
      "grad_norm": 1.7802496290562957,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.7751,
      "step": 21
    },
    {
      "epoch": 0.6683544303797468,
      "grad_norm": 1.1842415090537146,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.7538,
      "step": 22
    },
    {
      "epoch": 0.6987341772151898,
      "grad_norm": 1.4960633258701592,
      "learning_rate": 2e-05,
      "loss": 0.7926,
      "step": 23
    },
    {
      "epoch": 0.7291139240506329,
      "grad_norm": 1.5097669451949176,
      "learning_rate": 1.999877856940653e-05,
      "loss": 0.7449,
      "step": 24
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.962899630478963,
      "learning_rate": 1.999511457600466e-05,
      "loss": 0.7511,
      "step": 25
    },
    {
      "epoch": 0.789873417721519,
      "grad_norm": 1.014842989967261,
      "learning_rate": 1.9989008914857115e-05,
      "loss": 0.7487,
      "step": 26
    },
    {
      "epoch": 0.8202531645569621,
      "grad_norm": 1.038996274360424,
      "learning_rate": 1.998046307749216e-05,
      "loss": 0.7559,
      "step": 27
    },
    {
      "epoch": 0.850632911392405,
      "grad_norm": 1.0877346288126502,
      "learning_rate": 1.9969479151539238e-05,
      "loss": 0.7547,
      "step": 28
    },
    {
      "epoch": 0.8810126582278481,
      "grad_norm": 1.0095565903485795,
      "learning_rate": 1.9956059820218982e-05,
      "loss": 0.7245,
      "step": 29
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.8861577488710487,
      "learning_rate": 1.9940208361687762e-05,
      "loss": 0.7755,
      "step": 30
    },
    {
      "epoch": 0.9417721518987342,
      "grad_norm": 0.900902410227904,
      "learning_rate": 1.9921928648236855e-05,
      "loss": 0.7523,
      "step": 31
    },
    {
      "epoch": 0.9721518987341772,
      "grad_norm": 0.8791841576284923,
      "learning_rate": 1.990122514534651e-05,
      "loss": 0.7479,
      "step": 32
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 1.0519429196421413,
      "learning_rate": 1.9878102910595097e-05,
      "loss": 0.7165,
      "step": 33
    },
    {
      "epoch": 1.040506329113924,
      "grad_norm": 0.8934914710658114,
      "learning_rate": 1.985256759242359e-05,
      "loss": 0.6802,
      "step": 34
    },
    {
      "epoch": 1.070886075949367,
      "grad_norm": 0.9841402105625747,
      "learning_rate": 1.982462542875576e-05,
      "loss": 0.6758,
      "step": 35
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.9189686793774443,
      "learning_rate": 1.979428324547432e-05,
      "loss": 0.6474,
      "step": 36
    },
    {
      "epoch": 1.1316455696202532,
      "grad_norm": 0.9896237468293002,
      "learning_rate": 1.9761548454753455e-05,
      "loss": 0.6195,
      "step": 37
    },
    {
      "epoch": 1.1620253164556962,
      "grad_norm": 0.9334380903155232,
      "learning_rate": 1.972642905324813e-05,
      "loss": 0.642,
      "step": 38
    },
    {
      "epoch": 1.1924050632911392,
      "grad_norm": 0.9422054110909854,
      "learning_rate": 1.9688933620140638e-05,
      "loss": 0.6422,
      "step": 39
    },
    {
      "epoch": 1.2227848101265824,
      "grad_norm": 0.8638015865986586,
      "learning_rate": 1.96490713150448e-05,
      "loss": 0.6305,
      "step": 40
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.9014991175585031,
      "learning_rate": 1.9606851875768404e-05,
      "loss": 0.6066,
      "step": 41
    },
    {
      "epoch": 1.2835443037974683,
      "grad_norm": 0.8038121284043575,
      "learning_rate": 1.956228561593441e-05,
      "loss": 0.6576,
      "step": 42
    },
    {
      "epoch": 1.3139240506329113,
      "grad_norm": 1.3120410017449198,
      "learning_rate": 1.9515383422461457e-05,
      "loss": 0.6284,
      "step": 43
    },
    {
      "epoch": 1.3443037974683545,
      "grad_norm": 0.8692441037155122,
      "learning_rate": 1.9466156752904344e-05,
      "loss": 0.6136,
      "step": 44
    },
    {
      "epoch": 1.3746835443037975,
      "grad_norm": 0.9220257138930098,
      "learning_rate": 1.9414617632655114e-05,
      "loss": 0.6287,
      "step": 45
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.7900670784950345,
      "learning_rate": 1.9360778652005416e-05,
      "loss": 0.6173,
      "step": 46
    },
    {
      "epoch": 1.4354430379746836,
      "grad_norm": 0.8198297960820874,
      "learning_rate": 1.9304652963070868e-05,
      "loss": 0.6042,
      "step": 47
    },
    {
      "epoch": 1.4658227848101266,
      "grad_norm": 0.84500991999617,
      "learning_rate": 1.9246254276578175e-05,
      "loss": 0.6035,
      "step": 48
    },
    {
      "epoch": 1.4962025316455696,
      "grad_norm": 0.831254434919479,
      "learning_rate": 1.9185596858515797e-05,
      "loss": 0.6018,
      "step": 49
    },
    {
      "epoch": 1.5265822784810128,
      "grad_norm": 0.7529607201738153,
      "learning_rate": 1.9122695526648968e-05,
      "loss": 0.6125,
      "step": 50
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.9292969465457087,
      "learning_rate": 1.905756564689991e-05,
      "loss": 0.6289,
      "step": 51
    },
    {
      "epoch": 1.5873417721518988,
      "grad_norm": 0.7972979849139701,
      "learning_rate": 1.8990223129594146e-05,
      "loss": 0.6133,
      "step": 52
    },
    {
      "epoch": 1.6177215189873417,
      "grad_norm": 0.8678265928251405,
      "learning_rate": 1.8920684425573865e-05,
      "loss": 0.6366,
      "step": 53
    },
    {
      "epoch": 1.6481012658227847,
      "grad_norm": 0.846391215030445,
      "learning_rate": 1.884896652217917e-05,
      "loss": 0.6458,
      "step": 54
    },
    {
      "epoch": 1.678481012658228,
      "grad_norm": 0.8776578698645945,
      "learning_rate": 1.877508693909831e-05,
      "loss": 0.6057,
      "step": 55
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.6976854478409449,
      "learning_rate": 1.8699063724087905e-05,
      "loss": 0.6215,
      "step": 56
    },
    {
      "epoch": 1.7392405063291139,
      "grad_norm": 0.9568563158471319,
      "learning_rate": 1.862091544856407e-05,
      "loss": 0.6306,
      "step": 57
    },
    {
      "epoch": 1.769620253164557,
      "grad_norm": 0.8015223747864999,
      "learning_rate": 1.854066120306571e-05,
      "loss": 0.6008,
      "step": 58
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.7183123847613103,
      "learning_rate": 1.8458320592590976e-05,
      "loss": 0.5997,
      "step": 59
    },
    {
      "epoch": 1.830379746835443,
      "grad_norm": 0.7238530484854606,
      "learning_rate": 1.837391373180801e-05,
      "loss": 0.613,
      "step": 60
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.8839562029790872,
      "learning_rate": 1.8287461240141217e-05,
      "loss": 0.6232,
      "step": 61
    },
    {
      "epoch": 1.891139240506329,
      "grad_norm": 0.7381987053798049,
      "learning_rate": 1.8198984236734246e-05,
      "loss": 0.6147,
      "step": 62
    },
    {
      "epoch": 1.9215189873417722,
      "grad_norm": 0.6860283448515635,
      "learning_rate": 1.8108504335290852e-05,
      "loss": 0.5824,
      "step": 63
    },
    {
      "epoch": 1.9518987341772152,
      "grad_norm": 0.7731811152531224,
      "learning_rate": 1.8016043638794975e-05,
      "loss": 0.5892,
      "step": 64
    },
    {
      "epoch": 1.9822784810126581,
      "grad_norm": 0.8902621923934971,
      "learning_rate": 1.7921624734111292e-05,
      "loss": 0.6164,
      "step": 65
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 0.9381661022219966,
      "learning_rate": 1.7825270686467567e-05,
      "loss": 0.5494,
      "step": 66
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.8313369309395601,
      "learning_rate": 1.7727005033820117e-05,
      "loss": 0.5065,
      "step": 67
    },
    {
      "epoch": 2.081012658227848,
      "grad_norm": 1.0909784218437952,
      "learning_rate": 1.762685178110382e-05,
      "loss": 0.4688,
      "step": 68
    },
    {
      "epoch": 2.1113924050632913,
      "grad_norm": 0.8102185985136526,
      "learning_rate": 1.752483539436807e-05,
      "loss": 0.4695,
      "step": 69
    },
    {
      "epoch": 2.141772151898734,
      "grad_norm": 0.8150461699622455,
      "learning_rate": 1.7420980794800013e-05,
      "loss": 0.4631,
      "step": 70
    },
    {
      "epoch": 2.1721518987341772,
      "grad_norm": 0.7366434509016765,
      "learning_rate": 1.731531335263669e-05,
      "loss": 0.4651,
      "step": 71
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.6438366711057529,
      "learning_rate": 1.720785888096743e-05,
      "loss": 0.4452,
      "step": 72
    },
    {
      "epoch": 2.232911392405063,
      "grad_norm": 0.7849783705848422,
      "learning_rate": 1.7098643629428035e-05,
      "loss": 0.4786,
      "step": 73
    },
    {
      "epoch": 2.2632911392405064,
      "grad_norm": 0.6890485823095356,
      "learning_rate": 1.698769427778842e-05,
      "loss": 0.465,
      "step": 74
    },
    {
      "epoch": 2.293670886075949,
      "grad_norm": 0.7795571855935616,
      "learning_rate": 1.687503792943506e-05,
      "loss": 0.445,
      "step": 75
    },
    {
      "epoch": 2.3240506329113924,
      "grad_norm": 0.7203973128631659,
      "learning_rate": 1.6760702104750046e-05,
      "loss": 0.469,
      "step": 76
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.7914443018908193,
      "learning_rate": 1.664471473438822e-05,
      "loss": 0.4644,
      "step": 77
    },
    {
      "epoch": 2.3848101265822783,
      "grad_norm": 0.7788847057361931,
      "learning_rate": 1.6527104152454096e-05,
      "loss": 0.4373,
      "step": 78
    },
    {
      "epoch": 2.4151898734177215,
      "grad_norm": 0.6282032809289748,
      "learning_rate": 1.6407899089580263e-05,
      "loss": 0.4591,
      "step": 79
    },
    {
      "epoch": 2.4455696202531647,
      "grad_norm": 0.7751991250463963,
      "learning_rate": 1.628712866590885e-05,
      "loss": 0.4297,
      "step": 80
    },
    {
      "epoch": 2.4759493670886075,
      "grad_norm": 0.6803570213417126,
      "learning_rate": 1.6164822383977912e-05,
      "loss": 0.4338,
      "step": 81
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.644157203873823,
      "learning_rate": 1.604101012151436e-05,
      "loss": 0.4566,
      "step": 82
    },
    {
      "epoch": 2.536708860759494,
      "grad_norm": 0.699609682692562,
      "learning_rate": 1.5915722124135227e-05,
      "loss": 0.4337,
      "step": 83
    },
    {
      "epoch": 2.5670886075949366,
      "grad_norm": 0.6274950731367955,
      "learning_rate": 1.5788988997959115e-05,
      "loss": 0.4342,
      "step": 84
    },
    {
      "epoch": 2.59746835443038,
      "grad_norm": 0.6785191087516351,
      "learning_rate": 1.5660841702129533e-05,
      "loss": 0.4521,
      "step": 85
    },
    {
      "epoch": 2.6278481012658226,
      "grad_norm": 0.6458270559050598,
      "learning_rate": 1.5531311541251995e-05,
      "loss": 0.429,
      "step": 86
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.5825312231638327,
      "learning_rate": 1.540043015774676e-05,
      "loss": 0.4261,
      "step": 87
    },
    {
      "epoch": 2.688607594936709,
      "grad_norm": 0.6018838268174339,
      "learning_rate": 1.5268229524119007e-05,
      "loss": 0.4719,
      "step": 88
    },
    {
      "epoch": 2.7189873417721517,
      "grad_norm": 0.6442322744207832,
      "learning_rate": 1.513474193514842e-05,
      "loss": 0.4397,
      "step": 89
    },
    {
      "epoch": 2.749367088607595,
      "grad_norm": 1.0364876787461956,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.4447,
      "step": 90
    },
    {
      "epoch": 2.779746835443038,
      "grad_norm": 0.7810282079692747,
      "learning_rate": 1.4864036634258112e-05,
      "loss": 0.4594,
      "step": 91
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.6287807387430605,
      "learning_rate": 1.4726885051885654e-05,
      "loss": 0.4493,
      "step": 92
    },
    {
      "epoch": 2.840506329113924,
      "grad_norm": 0.8628908774357499,
      "learning_rate": 1.4588578757110359e-05,
      "loss": 0.4563,
      "step": 93
    },
    {
      "epoch": 2.8708860759493673,
      "grad_norm": 0.6762860512537575,
      "learning_rate": 1.4449151536240167e-05,
      "loss": 0.4408,
      "step": 94
    },
    {
      "epoch": 2.90126582278481,
      "grad_norm": 0.7111451520262403,
      "learning_rate": 1.4308637449409705e-05,
      "loss": 0.4492,
      "step": 95
    },
    {
      "epoch": 2.9316455696202532,
      "grad_norm": 0.8540120415622079,
      "learning_rate": 1.4167070822259868e-05,
      "loss": 0.4575,
      "step": 96
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.6560445807413064,
      "learning_rate": 1.402448623755254e-05,
      "loss": 0.4507,
      "step": 97
    },
    {
      "epoch": 2.992405063291139,
      "grad_norm": 0.7348786615787521,
      "learning_rate": 1.3880918526722497e-05,
      "loss": 0.4541,
      "step": 98
    },
    {
      "epoch": 3.030379746835443,
      "grad_norm": 0.8829664917688399,
      "learning_rate": 1.3736402761368597e-05,
      "loss": 0.3649,
      "step": 99
    },
    {
      "epoch": 3.060759493670886,
      "grad_norm": 0.6187698173591839,
      "learning_rate": 1.3590974244686248e-05,
      "loss": 0.3467,
      "step": 100
    },
    {
      "epoch": 3.091139240506329,
      "grad_norm": 1.0636943838369806,
      "learning_rate": 1.344466850284333e-05,
      "loss": 0.3503,
      "step": 101
    },
    {
      "epoch": 3.1215189873417724,
      "grad_norm": 0.8644487563619947,
      "learning_rate": 1.3297521276301666e-05,
      "loss": 0.3257,
      "step": 102
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.7089065387489814,
      "learning_rate": 1.3149568511086104e-05,
      "loss": 0.3403,
      "step": 103
    },
    {
      "epoch": 3.1822784810126583,
      "grad_norm": 0.816509013413704,
      "learning_rate": 1.300084635000341e-05,
      "loss": 0.3156,
      "step": 104
    },
    {
      "epoch": 3.212658227848101,
      "grad_norm": 1.3537144078088974,
      "learning_rate": 1.2851391123813075e-05,
      "loss": 0.3452,
      "step": 105
    },
    {
      "epoch": 3.2430379746835443,
      "grad_norm": 0.6532502720165345,
      "learning_rate": 1.2701239342352223e-05,
      "loss": 0.3121,
      "step": 106
    },
    {
      "epoch": 3.2734177215189875,
      "grad_norm": 0.6740921801165518,
      "learning_rate": 1.2550427685616767e-05,
      "loss": 0.3398,
      "step": 107
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.6775253664355564,
      "learning_rate": 1.239899299480098e-05,
      "loss": 0.3349,
      "step": 108
    },
    {
      "epoch": 3.3341772151898734,
      "grad_norm": 0.6070119488191431,
      "learning_rate": 1.2246972263297718e-05,
      "loss": 0.3356,
      "step": 109
    },
    {
      "epoch": 3.3645569620253166,
      "grad_norm": 0.6183068551734306,
      "learning_rate": 1.2094402627661447e-05,
      "loss": 0.3362,
      "step": 110
    },
    {
      "epoch": 3.3949367088607594,
      "grad_norm": 0.5727032944622328,
      "learning_rate": 1.1941321358536278e-05,
      "loss": 0.335,
      "step": 111
    },
    {
      "epoch": 3.4253164556962026,
      "grad_norm": 0.545216388664766,
      "learning_rate": 1.1787765851551296e-05,
      "loss": 0.3568,
      "step": 112
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.5917822422602147,
      "learning_rate": 1.1633773618185302e-05,
      "loss": 0.309,
      "step": 113
    },
    {
      "epoch": 3.4860759493670885,
      "grad_norm": 0.5536891349857951,
      "learning_rate": 1.14793822766033e-05,
      "loss": 0.328,
      "step": 114
    },
    {
      "epoch": 3.5164556962025317,
      "grad_norm": 0.586578773838644,
      "learning_rate": 1.132462954246688e-05,
      "loss": 0.3297,
      "step": 115
    },
    {
      "epoch": 3.546835443037975,
      "grad_norm": 0.5883860411935965,
      "learning_rate": 1.1169553219720828e-05,
      "loss": 0.3572,
      "step": 116
    },
    {
      "epoch": 3.5772151898734177,
      "grad_norm": 0.5265642178366109,
      "learning_rate": 1.1014191191358118e-05,
      "loss": 0.349,
      "step": 117
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.5365927907732009,
      "learning_rate": 1.085858141016566e-05,
      "loss": 0.3395,
      "step": 118
    },
    {
      "epoch": 3.6379746835443036,
      "grad_norm": 0.5218576944631613,
      "learning_rate": 1.070276188945293e-05,
      "loss": 0.3141,
      "step": 119
    },
    {
      "epoch": 3.668354430379747,
      "grad_norm": 0.5519668362160018,
      "learning_rate": 1.0546770693765859e-05,
      "loss": 0.3213,
      "step": 120
    },
    {
      "epoch": 3.69873417721519,
      "grad_norm": 0.5598460421391935,
      "learning_rate": 1.0390645929588197e-05,
      "loss": 0.3282,
      "step": 121
    },
    {
      "epoch": 3.729113924050633,
      "grad_norm": 0.572924986843309,
      "learning_rate": 1.0234425736032607e-05,
      "loss": 0.3617,
      "step": 122
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.5606413379105347,
      "learning_rate": 1.007814827552384e-05,
      "loss": 0.3374,
      "step": 123
    },
    {
      "epoch": 3.7898734177215188,
      "grad_norm": 0.5251980197872343,
      "learning_rate": 9.92185172447616e-06,
      "loss": 0.3349,
      "step": 124
    },
    {
      "epoch": 3.820253164556962,
      "grad_norm": 0.5358363108925137,
      "learning_rate": 9.765574263967397e-06,
      "loss": 0.3243,
      "step": 125
    },
    {
      "epoch": 3.850632911392405,
      "grad_norm": 0.5808885739154309,
      "learning_rate": 9.609354070411807e-06,
      "loss": 0.3288,
      "step": 126
    },
    {
      "epoch": 3.8810126582278484,
      "grad_norm": 0.5168986173736445,
      "learning_rate": 9.453229306234143e-06,
      "loss": 0.324,
      "step": 127
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.5004669366262062,
      "learning_rate": 9.297238110547075e-06,
      "loss": 0.3489,
      "step": 128
    },
    {
      "epoch": 3.9417721518987343,
      "grad_norm": 0.5654350075684064,
      "learning_rate": 9.14141858983434e-06,
      "loss": 0.3318,
      "step": 129
    },
    {
      "epoch": 3.972151898734177,
      "grad_norm": 0.5046554162412703,
      "learning_rate": 8.985808808641883e-06,
      "loss": 0.292,
      "step": 130
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 0.5119577140725842,
      "learning_rate": 8.830446780279175e-06,
      "loss": 0.3146,
      "step": 131
    },
    {
      "epoch": 4.040506329113924,
      "grad_norm": 0.6359795366548766,
      "learning_rate": 8.675370457533122e-06,
      "loss": 0.2536,
      "step": 132
    },
    {
      "epoch": 4.0708860759493675,
      "grad_norm": 0.42996729243235526,
      "learning_rate": 8.520617723396702e-06,
      "loss": 0.2427,
      "step": 133
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.4624884430415813,
      "learning_rate": 8.366226381814698e-06,
      "loss": 0.2329,
      "step": 134
    },
    {
      "epoch": 4.131645569620253,
      "grad_norm": 0.7471790953768577,
      "learning_rate": 8.212234148448708e-06,
      "loss": 0.256,
      "step": 135
    },
    {
      "epoch": 4.162025316455696,
      "grad_norm": 0.5121763704104856,
      "learning_rate": 8.058678641463724e-06,
      "loss": 0.2361,
      "step": 136
    },
    {
      "epoch": 4.192405063291139,
      "grad_norm": 0.44917466494575753,
      "learning_rate": 7.905597372338558e-06,
      "loss": 0.281,
      "step": 137
    },
    {
      "epoch": 4.222784810126583,
      "grad_norm": 0.5356612822589712,
      "learning_rate": 7.753027736702283e-06,
      "loss": 0.2608,
      "step": 138
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.49982499555676657,
      "learning_rate": 7.601007005199022e-06,
      "loss": 0.2615,
      "step": 139
    },
    {
      "epoch": 4.283544303797468,
      "grad_norm": 0.417797620610829,
      "learning_rate": 7.449572314383237e-06,
      "loss": 0.2653,
      "step": 140
    },
    {
      "epoch": 4.313924050632911,
      "grad_norm": 0.48456941397864417,
      "learning_rate": 7.298760657647779e-06,
      "loss": 0.2471,
      "step": 141
    },
    {
      "epoch": 4.3443037974683545,
      "grad_norm": 0.5134473818568167,
      "learning_rate": 7.148608876186931e-06,
      "loss": 0.2269,
      "step": 142
    },
    {
      "epoch": 4.374683544303798,
      "grad_norm": 0.47151673638601854,
      "learning_rate": 6.999153649996595e-06,
      "loss": 0.2468,
      "step": 143
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.44294094632402514,
      "learning_rate": 6.8504314889138956e-06,
      "loss": 0.2767,
      "step": 144
    },
    {
      "epoch": 4.435443037974683,
      "grad_norm": 0.4397883834687915,
      "learning_rate": 6.702478723698336e-06,
      "loss": 0.2421,
      "step": 145
    },
    {
      "epoch": 4.465822784810126,
      "grad_norm": 0.46890058728506934,
      "learning_rate": 6.555331497156671e-06,
      "loss": 0.2543,
      "step": 146
    },
    {
      "epoch": 4.49620253164557,
      "grad_norm": 0.4572739193799271,
      "learning_rate": 6.4090257553137566e-06,
      "loss": 0.2577,
      "step": 147
    },
    {
      "epoch": 4.526582278481013,
      "grad_norm": 0.42261050259743843,
      "learning_rate": 6.263597238631405e-06,
      "loss": 0.2941,
      "step": 148
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.44366811059182965,
      "learning_rate": 6.119081473277502e-06,
      "loss": 0.2672,
      "step": 149
    },
    {
      "epoch": 4.587341772151898,
      "grad_norm": 0.42344675637556384,
      "learning_rate": 5.975513762447465e-06,
      "loss": 0.2417,
      "step": 150
    },
    {
      "epoch": 4.6177215189873415,
      "grad_norm": 0.4508361113487649,
      "learning_rate": 5.832929177740134e-06,
      "loss": 0.263,
      "step": 151
    },
    {
      "epoch": 4.648101265822785,
      "grad_norm": 0.4686806977031232,
      "learning_rate": 5.6913625505902966e-06,
      "loss": 0.2346,
      "step": 152
    },
    {
      "epoch": 4.678481012658228,
      "grad_norm": 0.41473474647323366,
      "learning_rate": 5.550848463759835e-06,
      "loss": 0.2784,
      "step": 153
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.4505275110926776,
      "learning_rate": 5.411421242889643e-06,
      "loss": 0.1977,
      "step": 154
    },
    {
      "epoch": 4.739240506329114,
      "grad_norm": 0.4205845928908748,
      "learning_rate": 5.273114948114346e-06,
      "loss": 0.2375,
      "step": 155
    },
    {
      "epoch": 4.769620253164557,
      "grad_norm": 0.4101090471186687,
      "learning_rate": 5.135963365741892e-06,
      "loss": 0.2421,
      "step": 156
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.42401339742540434,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.2752,
      "step": 157
    },
    {
      "epoch": 4.830379746835443,
      "grad_norm": 0.43759520975785615,
      "learning_rate": 4.865258064851579e-06,
      "loss": 0.2497,
      "step": 158
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.40500636421053166,
      "learning_rate": 4.731770475880995e-06,
      "loss": 0.2636,
      "step": 159
    },
    {
      "epoch": 4.891139240506329,
      "grad_norm": 0.41316294123941627,
      "learning_rate": 4.599569842253244e-06,
      "loss": 0.251,
      "step": 160
    },
    {
      "epoch": 4.921518987341772,
      "grad_norm": 0.4086719758074519,
      "learning_rate": 4.468688458748006e-06,
      "loss": 0.2484,
      "step": 161
    },
    {
      "epoch": 4.951898734177215,
      "grad_norm": 0.4188727111744998,
      "learning_rate": 4.339158297870469e-06,
      "loss": 0.2322,
      "step": 162
    },
    {
      "epoch": 4.982278481012658,
      "grad_norm": 0.8625576785136497,
      "learning_rate": 4.211011002040885e-06,
      "loss": 0.2361,
      "step": 163
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 0.5036794714725785,
      "learning_rate": 4.084277875864776e-06,
      "loss": 0.2212,
      "step": 164
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.4944010941746153,
      "learning_rate": 3.958989878485644e-06,
      "loss": 0.2182,
      "step": 165
    },
    {
      "epoch": 5.0810126582278485,
      "grad_norm": 0.34874875231904934,
      "learning_rate": 3.83517761602209e-06,
      "loss": 0.2238,
      "step": 166
    },
    {
      "epoch": 5.111392405063291,
      "grad_norm": 0.40123635946627545,
      "learning_rate": 3.712871334091154e-06,
      "loss": 0.2253,
      "step": 167
    },
    {
      "epoch": 5.141772151898734,
      "grad_norm": 0.4415603560120399,
      "learning_rate": 3.592100910419738e-06,
      "loss": 0.2006,
      "step": 168
    },
    {
      "epoch": 5.172151898734177,
      "grad_norm": 0.5082587245608364,
      "learning_rate": 3.4728958475459052e-06,
      "loss": 0.1923,
      "step": 169
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.46528984560958797,
      "learning_rate": 3.355285265611784e-06,
      "loss": 0.2249,
      "step": 170
    },
    {
      "epoch": 5.232911392405064,
      "grad_norm": 0.37638887667716,
      "learning_rate": 3.2392978952499553e-06,
      "loss": 0.1925,
      "step": 171
    },
    {
      "epoch": 5.263291139240506,
      "grad_norm": 0.3740637415615642,
      "learning_rate": 3.1249620705649417e-06,
      "loss": 0.206,
      "step": 172
    },
    {
      "epoch": 5.293670886075949,
      "grad_norm": 0.4642841524709166,
      "learning_rate": 3.0123057222115835e-06,
      "loss": 0.2133,
      "step": 173
    },
    {
      "epoch": 5.324050632911392,
      "grad_norm": 0.40642794096271934,
      "learning_rate": 2.9013563705719673e-06,
      "loss": 0.1961,
      "step": 174
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.37841545442942737,
      "learning_rate": 2.7921411190325753e-06,
      "loss": 0.2057,
      "step": 175
    },
    {
      "epoch": 5.384810126582279,
      "grad_norm": 0.33871095044107963,
      "learning_rate": 2.6846866473633126e-06,
      "loss": 0.1989,
      "step": 176
    },
    {
      "epoch": 5.415189873417722,
      "grad_norm": 0.3754823759393782,
      "learning_rate": 2.579019205199992e-06,
      "loss": 0.2126,
      "step": 177
    },
    {
      "epoch": 5.445569620253164,
      "grad_norm": 0.3856466030246529,
      "learning_rate": 2.4751646056319334e-06,
      "loss": 0.2257,
      "step": 178
    },
    {
      "epoch": 5.4759493670886075,
      "grad_norm": 0.3923098862110596,
      "learning_rate": 2.373148218896182e-06,
      "loss": 0.2144,
      "step": 179
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.37176203527945706,
      "learning_rate": 2.2729949661798876e-06,
      "loss": 0.2233,
      "step": 180
    },
    {
      "epoch": 5.536708860759494,
      "grad_norm": 0.35608814977164444,
      "learning_rate": 2.174729313532433e-06,
      "loss": 0.1819,
      "step": 181
    },
    {
      "epoch": 5.567088607594937,
      "grad_norm": 0.3243936037942546,
      "learning_rate": 2.078375265888707e-06,
      "loss": 0.1827,
      "step": 182
    },
    {
      "epoch": 5.597468354430379,
      "grad_norm": 0.33701700084967295,
      "learning_rate": 1.9839563612050273e-06,
      "loss": 0.2104,
      "step": 183
    },
    {
      "epoch": 5.627848101265823,
      "grad_norm": 0.33662685029898104,
      "learning_rate": 1.8914956647091497e-06,
      "loss": 0.1922,
      "step": 184
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.3473768780477551,
      "learning_rate": 1.8010157632657544e-06,
      "loss": 0.2041,
      "step": 185
    },
    {
      "epoch": 5.688607594936709,
      "grad_norm": 0.3626330297915193,
      "learning_rate": 1.7125387598587862e-06,
      "loss": 0.1918,
      "step": 186
    },
    {
      "epoch": 5.718987341772152,
      "grad_norm": 0.321473962236254,
      "learning_rate": 1.6260862681919965e-06,
      "loss": 0.2043,
      "step": 187
    },
    {
      "epoch": 5.749367088607595,
      "grad_norm": 0.35704600426009947,
      "learning_rate": 1.5416794074090258e-06,
      "loss": 0.2017,
      "step": 188
    },
    {
      "epoch": 5.779746835443038,
      "grad_norm": 0.3947516623595968,
      "learning_rate": 1.459338796934293e-06,
      "loss": 0.1679,
      "step": 189
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.3474014346693818,
      "learning_rate": 1.3790845514359363e-06,
      "loss": 0.2256,
      "step": 190
    },
    {
      "epoch": 5.840506329113924,
      "grad_norm": 0.32385363472954287,
      "learning_rate": 1.300936275912098e-06,
      "loss": 0.2154,
      "step": 191
    },
    {
      "epoch": 5.870886075949367,
      "grad_norm": 0.33036802999279485,
      "learning_rate": 1.224913060901688e-06,
      "loss": 0.2126,
      "step": 192
    },
    {
      "epoch": 5.9012658227848105,
      "grad_norm": 0.3394966568871476,
      "learning_rate": 1.1510334778208332e-06,
      "loss": 0.1767,
      "step": 193
    },
    {
      "epoch": 5.931645569620253,
      "grad_norm": 0.3277144106012969,
      "learning_rate": 1.0793155744261352e-06,
      "loss": 0.1926,
      "step": 194
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.3069253295356838,
      "learning_rate": 1.0097768704058542e-06,
      "loss": 0.1864,
      "step": 195
    },
    {
      "epoch": 5.992405063291139,
      "grad_norm": 0.3255939897446572,
      "learning_rate": 9.424343531000968e-07,
      "loss": 0.2161,
      "step": 196
    },
    {
      "epoch": 6.030379746835443,
      "grad_norm": 0.38232908425312045,
      "learning_rate": 8.773044733510338e-07,
      "loss": 0.2003,
      "step": 197
    },
    {
      "epoch": 6.060759493670886,
      "grad_norm": 0.33859768140109414,
      "learning_rate": 8.144031414842012e-07,
      "loss": 0.1646,
      "step": 198
    },
    {
      "epoch": 6.091139240506329,
      "grad_norm": 0.32706697650560945,
      "learning_rate": 7.537457234218271e-07,
      "loss": 0.2051,
      "step": 199
    },
    {
      "epoch": 6.121518987341772,
      "grad_norm": 0.27896217677512697,
      "learning_rate": 6.953470369291349e-07,
      "loss": 0.1838,
      "step": 200
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.31483538844232006,
      "learning_rate": 6.392213479945852e-07,
      "loss": 0.1738,
      "step": 201
    },
    {
      "epoch": 6.182278481012658,
      "grad_norm": 0.2765477473961978,
      "learning_rate": 5.853823673448877e-07,
      "loss": 0.1828,
      "step": 202
    },
    {
      "epoch": 6.2126582278481015,
      "grad_norm": 0.2936300337575462,
      "learning_rate": 5.33843247095659e-07,
      "loss": 0.2009,
      "step": 203
    },
    {
      "epoch": 6.243037974683545,
      "grad_norm": 0.29287413895540865,
      "learning_rate": 4.846165775385459e-07,
      "loss": 0.1801,
      "step": 204
    },
    {
      "epoch": 6.273417721518987,
      "grad_norm": 0.2962493920168721,
      "learning_rate": 4.3771438406559173e-07,
      "loss": 0.1785,
      "step": 205
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.3164172270109455,
      "learning_rate": 3.931481242315993e-07,
      "loss": 0.1885,
      "step": 206
    },
    {
      "epoch": 6.334177215189873,
      "grad_norm": 0.29197813417087864,
      "learning_rate": 3.5092868495520294e-07,
      "loss": 0.2003,
      "step": 207
    },
    {
      "epoch": 6.364556962025317,
      "grad_norm": 0.31929773002115464,
      "learning_rate": 3.110663798593616e-07,
      "loss": 0.1927,
      "step": 208
    },
    {
      "epoch": 6.39493670886076,
      "grad_norm": 0.29045637553524123,
      "learning_rate": 2.735709467518699e-07,
      "loss": 0.1698,
      "step": 209
    },
    {
      "epoch": 6.425316455696202,
      "grad_norm": 0.30725066330109196,
      "learning_rate": 2.384515452465475e-07,
      "loss": 0.1928,
      "step": 210
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.30615861813107187,
      "learning_rate": 2.0571675452567997e-07,
      "loss": 0.1782,
      "step": 211
    },
    {
      "epoch": 6.4860759493670885,
      "grad_norm": 0.2869563852197699,
      "learning_rate": 1.7537457124423896e-07,
      "loss": 0.1623,
      "step": 212
    },
    {
      "epoch": 6.516455696202532,
      "grad_norm": 0.30672424574309864,
      "learning_rate": 1.474324075764111e-07,
      "loss": 0.1847,
      "step": 213
    },
    {
      "epoch": 6.546835443037975,
      "grad_norm": 0.3101103976254337,
      "learning_rate": 1.2189708940490653e-07,
      "loss": 0.192,
      "step": 214
    },
    {
      "epoch": 6.577215189873417,
      "grad_norm": 0.2980650463814902,
      "learning_rate": 9.877485465349057e-08,
      "loss": 0.2099,
      "step": 215
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.2990087383345418,
      "learning_rate": 7.807135176314707e-08,
      "loss": 0.2154,
      "step": 216
    },
    {
      "epoch": 6.637974683544304,
      "grad_norm": 0.2982897237752021,
      "learning_rate": 5.979163831223988e-08,
      "loss": 0.1959,
      "step": 217
    },
    {
      "epoch": 6.668354430379747,
      "grad_norm": 0.2978342853151662,
      "learning_rate": 4.394017978101905e-08,
      "loss": 0.2098,
      "step": 218
    },
    {
      "epoch": 6.69873417721519,
      "grad_norm": 0.2811890435403091,
      "learning_rate": 3.0520848460765525e-08,
      "loss": 0.179,
      "step": 219
    },
    {
      "epoch": 6.729113924050633,
      "grad_norm": 0.3155274310838752,
      "learning_rate": 1.9536922507841227e-08,
      "loss": 0.1973,
      "step": 220
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.30061370543214616,
      "learning_rate": 1.099108514288627e-08,
      "loss": 0.171,
      "step": 221
    },
    {
      "epoch": 6.789873417721519,
      "grad_norm": 0.2857679692734632,
      "learning_rate": 4.885423995341088e-09,
      "loss": 0.1863,
      "step": 222
    },
    {
      "epoch": 6.820253164556962,
      "grad_norm": 0.29067938294142787,
      "learning_rate": 1.2214305934699078e-09,
      "loss": 0.1882,
      "step": 223
    },
    {
      "epoch": 6.850632911392405,
      "grad_norm": 0.3036179454626494,
      "learning_rate": 0.0,
      "loss": 0.1986,
      "step": 224
    },
    {
      "epoch": 6.850632911392405,
      "step": 224,
      "total_flos": 3.784673209782108e+17,
      "train_loss": 0.421367068442383,
      "train_runtime": 25625.0045,
      "train_samples_per_second": 0.863,
      "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": 3.784673209782108e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}