{
  "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": 6.985423929375897,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.9457,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 7.635900060835987,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0165,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 6.925059850892843,
      "learning_rate": 2.5e-06,
      "loss": 1.0366,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 7.5923027114852495,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0122,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 5.174846166818154,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9254,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 4.1098376697450325,
      "learning_rate": 5e-06,
      "loss": 0.9647,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 3.393317627304308,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.914,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 3.0110605381333473,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.8691,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 3.1971822819519424,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.8462,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 2.7994264065082968,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.7583,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 2.97745465815888,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.7174,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 3.2909313513400247,
      "learning_rate": 1e-05,
      "loss": 0.8281,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 2.4968336927762174,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.7152,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 1.8662888941616411,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.6728,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 2.264821943619885,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.7307,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 1.8660773999817588,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.6258,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 2.074162044859812,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.6115,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 1.9536447272103374,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.6621,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 1.7565106269565998,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.6426,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.2140295341842848,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.5377,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 1.4173797019125525,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.6157,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 1.248309422022285,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.5267,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 1.1969476726468316,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.5002,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 1.3439477608044215,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.4973,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 1.3512891206720117,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.5686,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 1.3532231708677511,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.453,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 1.4547893582078886,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.5362,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 1.1035603719215918,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.4751,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 1.0577386601742778,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.4695,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 1.0263972961066887,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.4252,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 0.9171889641703226,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.397,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 0.8942651622425055,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.3685,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 1.4190230540501483,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.3762,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 0.9008345219541247,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.4226,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 1.115950905448313,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.3912,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 1.3107407416164085,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.3695,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 1.0780031944765727,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.3987,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 0.9732979125485598,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.3722,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.7485040299560034,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.3469,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.9068526312087641,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.3401,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 1.0352985676903166,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.367,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.8384007166279289,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.2696,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 0.8668157138858013,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.3087,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 0.9027643625440266,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.3239,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.9124051780232874,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.2656,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.9476745421093329,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.2963,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.9628110321431402,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3296,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.8813705327288355,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.2582,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.8047378895305227,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.3102,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.702713716125064,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.2222,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.7597146269306173,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.2291,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.8657916526771118,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.2867,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.9300384682568652,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.1868,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.8095892820761132,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.1927,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.8510881476139399,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.3279,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.9021622365538562,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.209,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.9725356334490954,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.2742,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.5645411887583268,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.2199,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.6518076848320741,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.1918,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.6377630336308364,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.197,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.6347394964635771,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.1614,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.8535536743198393,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.2126,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.9221943586925377,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.1935,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.6792159680683504,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.1897,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.8373189417298913,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.2797,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.8971515546776333,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.2029,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 0.6640164305867826,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.1774,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 0.47299617494638746,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.1827,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.6014815198419162,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.1929,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 0.6185839486947972,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.1627,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 0.6123534348227857,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.2197,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.4011381343318599,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.0725,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.5722110429360766,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.2184,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.6671983375782435,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.2178,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.5868743301774155,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.1916,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 0.8187021507762621,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.1362,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.3670283736721863,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.1607,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.5332992797967622,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.1786,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.4708383681991265,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.1642,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.4944947758749545,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.1959,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.46495664786060314,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.1642,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.5630367423283471,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.1059,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.52535620826898,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.1773,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.47731703493252253,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.0969,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.6563447114396115,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.1306,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 0.3933039934976602,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.1441,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.5033551982043885,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.1396,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.39618402225000293,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.0744,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.4565754807129156,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.1932,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 0.43244140513240303,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.2042,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.4401904045308233,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.1034,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.4202092979555295,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.1635,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.3991721698118227,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.1487,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.3207727357885671,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.0803,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 0.4739109244788485,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.1647,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 0.348232125198483,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.1417,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.3243625959958066,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.1102,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.31945251529448787,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.1051,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.3712326109174239,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.2026,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.34573823680285704,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.1969,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.29996869319971203,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.0823,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.22257299251251544,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.0915,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.34683650982572856,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.1517,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.3105679651402835,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.1393,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.3028763915701349,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.1027,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.2799516037692743,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.1036,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.2894370705831595,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.1259,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.27708635216418004,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.0926,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.32711320555461687,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.1267,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.34493716708207217,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.1769,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.26915554056975277,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.1202,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.326945137717795,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.1135,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.35216083955448363,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.1349,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.3599202068234174,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.187,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.2282614434684352,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.0426,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.3167109588751429,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.1372,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.3531077213910155,
      "learning_rate": 0.0,
      "loss": 0.1904,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 5.43823155429376e+16,
      "train_loss": 0.0,
      "train_runtime": 2.2596,
      "train_samples_per_second": 1817.981,
      "train_steps_per_second": 51.778
    }
  ],
  "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": 5.43823155429376e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}