{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 12.30379746835443,
  "eval_steps": 500,
  "global_step": 117,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 6.157211549281161,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.0626,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 6.804001217549786,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.2248,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 6.597084767459683,
      "learning_rate": 2.5e-06,
      "loss": 1.1052,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 5.942177286597682,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.1595,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 5.418460452000427,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.1239,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 3.153293687100172,
      "learning_rate": 5e-06,
      "loss": 1.0674,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 2.806566824802429,
      "learning_rate": 5.833333333333334e-06,
      "loss": 1.0055,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 3.338941106153212,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9936,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 3.302474540308339,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9816,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 3.595513695722795,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9489,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 3.3871964531791683,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.9274,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 3.0857414150366087,
      "learning_rate": 1e-05,
      "loss": 0.9095,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 2.5191620958567182,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.8981,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 2.737738731566357,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.8433,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 2.0834015709469913,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.8093,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 2.0416887226196576,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.7986,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 1.8246193094084553,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.8024,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 1.9375402731950186,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.8504,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 1.6580702258728945,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.7831,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.4278347437478978,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.7317,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 1.320789126847037,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.7074,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 1.3812636861307406,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.6884,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 1.2134556666293617,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.6394,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 1.079665738094554,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.6071,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 1.1043875436911461,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.6714,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 1.2548432027386867,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.5986,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 1.185359569323539,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.6327,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 1.0349166830436058,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.6285,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 1.0508235984952188,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.6241,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 1.0513318043556434,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.5083,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 0.8822283897826719,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.4989,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 1.041645503901323,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.5224,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 1.0004920789603473,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.5043,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 1.0583378433032402,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.5066,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 1.137619127899739,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.4884,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 1.1362086891510303,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.5393,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 1.0203109658125569,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.5067,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 1.018350096806046,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.4978,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.9783814641328696,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.4271,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.9836885405969577,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.4186,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 1.0414967810586182,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.4145,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.9533251563846102,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.3187,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 0.9491516155178044,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.4154,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 1.0480122702954697,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.405,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 1.07977730400388,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.3854,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.9520505922598491,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.3138,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.9342729262774867,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3596,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.7890396667618882,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.3498,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.6769564798889887,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.3278,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.7653168892895088,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.3213,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.821437757032844,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.3747,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.8065078587809685,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.3109,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.8283436174119972,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.2898,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.7133401437407201,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.2991,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.8393820200936759,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.2762,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.8080598708894239,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.2439,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.7541318136734192,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.3137,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.6766546555645336,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.2574,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.5352833431889403,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.1956,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.724724143088227,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.2565,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.8806497312385178,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.3037,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.9776757860743662,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.2369,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.783780176134182,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.1729,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 1.0013521096267555,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.2579,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.9190562045062212,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.2782,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.7985078967653398,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.2408,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 0.7061403046585132,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.2107,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 0.5353660835391822,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.2229,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.5349960185147552,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.1277,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 0.8664628478735551,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.2897,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 0.8253839737808106,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.2406,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.7421630688711993,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.1761,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.5837425897812744,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.1929,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.6737135323571432,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.1819,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.824546239321345,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.2077,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 0.6304741700396211,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.1836,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.565617840655265,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.1977,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.5339151664767571,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.158,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.47880796619284083,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.1434,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.49300724389920564,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.1397,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.7367074010978124,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.2219,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.7935917293523205,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.1362,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.81357322148699,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.1965,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.6783739078266938,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.2106,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.6775925302551934,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.151,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 0.6052272364557885,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.1684,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.68487915195375,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.1226,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.6459472434105652,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.1742,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.474462925301277,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.1384,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 0.4152229891452817,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.1457,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.4946936552850661,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.1634,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.5072275236865679,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.1603,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.5742269430860725,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.1641,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.6074824362819786,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.2185,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 0.6793854820235138,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.131,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 0.40893746572038137,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.1025,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.5342454310878073,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.1279,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.43604097423308696,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.1781,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.3928339993546957,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.1564,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.4479932860670764,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.1901,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.3796740429736731,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.1512,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.34531475573057335,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.1403,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.4170699700689817,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.1152,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.3748966707364169,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.1505,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.32636197875820894,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.1175,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.38481647415256054,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.1952,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.3733729710238798,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.1896,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.32423284456950824,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.1175,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.40945229473076095,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.1791,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.30144146921476134,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.0965,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.3278332838255955,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.0865,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.33444975973059593,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.1094,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.41469846871433114,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.1841,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.38577746843512856,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.1457,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.3447840296489404,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.1491,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.3838210348234713,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.1698,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.36652031235159765,
      "learning_rate": 0.0,
      "loss": 0.1431,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 6.615101660043674e+16,
      "train_loss": 0.3977203600936466,
      "train_runtime": 3641.2242,
      "train_samples_per_second": 1.128,
      "train_steps_per_second": 0.032
    }
  ],
  "logging_steps": 1,
  "max_steps": 117,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 13,
  "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": 6.615101660043674e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}