{
  "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": 6.860558090768261,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.0961,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 6.6465514328000745,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0624,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 6.7336591180215,
      "learning_rate": 2.5e-06,
      "loss": 1.0408,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 6.289410141432145,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.1017,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 5.297067935553607,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.0212,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 2.8715507373739246,
      "learning_rate": 5e-06,
      "loss": 1.0147,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 2.4840040013274916,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9425,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 3.9751721501920887,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9312,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 4.580762956977892,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9666,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 4.925615088993872,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.0278,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 4.233564087590778,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.9161,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 4.001833935766076,
      "learning_rate": 1e-05,
      "loss": 0.8719,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 3.4245623695907597,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.8726,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 2.696132108558983,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.8577,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 2.31300763626559,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.7491,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 2.3234586548237903,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.7734,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 2.210462085807912,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.7982,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 2.090064765127792,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.7956,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 1.6498995178157663,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.7598,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.3260632500193477,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.7224,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 2.0177969688243875,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.6429,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 1.3947855309290407,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.6298,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 1.160738626144563,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.6423,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.9611984492228803,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.6368,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 1.2938115981966862,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.6637,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 1.2964949730488236,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.5964,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 1.2803815585752587,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.6773,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 1.0417027540888173,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.5915,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 1.0612408355489449,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.5368,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 1.015385383584571,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.5076,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 0.9364357273481981,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.4546,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 1.2102775424844903,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.5021,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.9436194933396135,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.4641,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 0.9236675911494436,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.5377,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 0.9701997512671356,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.4682,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.966002707537885,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.4838,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 0.9673557353484119,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.4082,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 1.034108411882538,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.486,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.9415810709026631,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.376,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.823044705358147,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.3811,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 0.9145320708132261,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.3526,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.8650092747944037,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.328,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 0.9881756344685371,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.349,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 0.9823509382798025,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.3598,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.8958935466578937,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.2877,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.9118557340368669,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.311,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.8763927937606902,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3325,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.7461072521001721,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.3349,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.7642790021862156,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.2807,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.7975056179101765,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.2098,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.9124369831781823,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.2594,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.9492360805787992,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.2787,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.7810133018188129,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.2579,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.797261311243426,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.254,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.8256511482694303,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.2,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.8368402309835822,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.242,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.9079303016160376,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.289,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.7062156273760553,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.1659,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.6364183633169183,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.1812,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.6485426859798974,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.191,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.7343183475245836,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.1901,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.8842239669919117,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.1686,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.7652713087618732,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.2216,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.7476641703617927,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.2111,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.7908250351536755,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.1748,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.8028977222691924,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.1742,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 0.669111102115808,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.1885,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 0.6513520605082463,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.1127,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.7284123946336519,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.157,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 0.9141575301556538,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.169,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 0.6627033047344973,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.1424,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.6118704094500598,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.17,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.7477772617221755,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.1254,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.6253841123681729,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.1452,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.7355131370953542,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.1519,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 0.5744600976061411,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.1336,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.4461430786464735,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.1052,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.4926864535057879,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.1154,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.5755178947690937,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.1642,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.6275168864705533,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.1304,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.6186003802401066,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.1055,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.5086223940714127,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.1122,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.5565024636347086,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.1169,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.47674281005279834,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.0963,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.5071656966074303,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.1181,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 0.5080428479009413,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.0986,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.5767610963915656,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.0749,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.5802766101644845,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.1261,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.36483204283162396,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.0864,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 0.3484838531621286,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.0866,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.4207978870828214,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.0867,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.36609838515178533,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.0912,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.4041800913892658,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.0963,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.4737359478690017,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.1091,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 0.511893786876196,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.1147,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 0.505741190251688,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.1419,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.2921652214779831,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.0622,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.3206284710439013,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.0911,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.3101356512139978,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.0571,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.4170506484120927,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.1208,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.30570976039907305,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.0539,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.36911481022909254,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.0951,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.32390715963815525,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.0888,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.2913073567786816,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.0652,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.3664972128112503,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.1228,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.34182608058603037,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.0948,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.3486549100985537,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.0687,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.3157350540509197,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.0917,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.36754592677523107,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.0745,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.39375574205698355,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.0805,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.30117953415296717,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.0855,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.3112510420055146,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.0697,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.33876280179133755,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.1091,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.28373207785945503,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.0769,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.33096901791472194,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.0921,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.31140721315301356,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.063,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.36234347986390786,
      "learning_rate": 0.0,
      "loss": 0.1417,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 6.172680040061338e+16,
      "train_loss": 0.3512128319941525,
      "train_runtime": 5828.0451,
      "train_samples_per_second": 0.705,
      "train_steps_per_second": 0.02
    }
  ],
  "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": 6.172680040061338e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}