{
  "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": 6.5003140058321405,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.8149,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 6.385237960408739,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.8266,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 6.268599125012116,
      "learning_rate": 2.5e-06,
      "loss": 0.8107,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 5.371843727828055,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.7504,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 4.090272173193176,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.7218,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 2.4352362376060985,
      "learning_rate": 5e-06,
      "loss": 0.7474,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 1.9633513732675176,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.7043,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 3.6769350435918713,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.7242,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 4.214439050914602,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7273,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 6.065015004552783,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.0803,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 3.913406195756194,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.7003,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 2.997646583503768,
      "learning_rate": 1e-05,
      "loss": 0.62,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 2.1072932466633736,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.679,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 1.5148197177552465,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.6227,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 1.758914261566612,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.6173,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 1.486593346759676,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.6245,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 1.0478636382609359,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.561,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.898661013744487,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.5571,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 1.7807186148797274,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.8308,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.04719076470331,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.5674,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.7825853813436014,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.5271,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 1.0428621059543883,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.5325,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 0.8154021057484789,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.5028,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.6934752614411399,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.5189,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 0.8048252618770353,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.5196,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 0.902568556163563,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.5051,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.7228722034288294,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.5171,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 0.6048464582031282,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.518,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 1.28108116328936,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.7108,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.6207656255691539,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.4261,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 0.7988771425464194,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.4849,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 0.7955102285343925,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.4583,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.6682056184909779,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.4355,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 0.6655553236668657,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.4931,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 0.6136027303515259,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.4542,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.6226581021953237,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.3953,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 0.6465762604855989,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.4483,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 0.8243420731830164,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.6539,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.5071957500121406,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.397,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.6062659619653404,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.3915,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 0.6273385685870636,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.3942,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.6085918174570976,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.4063,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 0.6253442911527379,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.3837,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 0.6474785505358827,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.4078,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.5994227242123352,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.3884,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.5619436467867244,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.3847,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.5349712880208912,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3589,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.9308206699510984,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.5596,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.5053193080773055,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.3441,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.5859856236988434,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.3327,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.5374680301200689,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.3449,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.6296934706514044,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.3497,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.5368080590736646,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.3277,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.49955903439548077,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.3145,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.5053002734143563,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.3076,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.5018251834226533,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.3173,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.8333804290377321,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.4729,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.5210323632451853,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.3053,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.4972963343442431,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.2939,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.5358246742798298,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.276,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.6032343128120313,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.2716,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.49005609210360823,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.2824,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.4783469654146685,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.2708,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.6271069482337951,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.2678,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.4703217117469394,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.2637,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.5742394904378342,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.2584,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 0.7866786244627546,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.3858,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 0.5339513380666302,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.238,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.5368504319625207,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.2568,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 0.48712805470619197,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.242,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 0.5757186722523749,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.2266,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.5319047801575709,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.2244,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.4422402128461875,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.2199,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.49172302256412104,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.2209,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.5209827613550508,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.2257,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 0.7111509875319255,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.3417,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.4557857288699775,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.2051,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.5000449002749855,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.1877,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.535621952261449,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.197,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.46642861373477995,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.2088,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.4615151283659087,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.1934,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.45458756404451356,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.1877,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.46583002632277315,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.2025,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.45074860231419794,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.1798,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.4673709474566254,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.1844,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 0.6274710894664994,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.2608,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.509643818370377,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.1727,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.4406801893716253,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.1636,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.4331285250869376,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.1816,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 0.46376899491296963,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.1683,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.5648932341873285,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.1661,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.4901807707527251,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.1718,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.41370464591210493,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.1567,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.4354455522212687,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.1856,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 0.7129152816721355,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.2311,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 0.49860684407976075,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.1607,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.4529491327532631,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.1536,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.4208802989135161,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.16,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.39036814021015803,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.1567,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.48017703008509716,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.1487,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.4487704987423807,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.155,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.4568400069361551,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.1514,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.46902530835895423,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.1536,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.41468068415296566,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.1551,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.5772963558578911,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.203,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.37643659865935974,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.1532,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.3841271773787093,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.1614,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.38500491197383674,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.1471,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.37112497055801563,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.1388,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.4082267597413116,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.1538,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.37239152460407554,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.1464,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.3538955499435395,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.1544,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.36891082481678095,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.1365,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.5711614644507576,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.2285,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.3464442741783666,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.1412,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.42114936930553215,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.1536,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.34626902368278806,
      "learning_rate": 0.0,
      "loss": 0.1454,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 1.3646752746622157e+17,
      "train_loss": 0.36502268630215245,
      "train_runtime": 5010.9477,
      "train_samples_per_second": 0.82,
      "train_steps_per_second": 0.023
    }
  ],
  "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": 1.3646752746622157e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}