{
  "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": 6.1036845704758855,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.7812,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 6.211931564456468,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.8133,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 5.790768954239727,
      "learning_rate": 2.5e-06,
      "loss": 0.7502,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 6.013932016991219,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.8266,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 4.390681107164878,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.7591,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 2.373460332124642,
      "learning_rate": 5e-06,
      "loss": 0.6998,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 2.049676391051526,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.6863,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 3.6997148412349437,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.6978,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 3.5256032920681037,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6739,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 6.2693023812601725,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.029,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 4.158105021945833,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.6936,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 3.520460730229181,
      "learning_rate": 1e-05,
      "loss": 0.6721,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 2.1407618632862238,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.6373,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 1.346669398553155,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.5256,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 1.83167377388535,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.6005,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 1.6292211200987232,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.6091,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 1.0642263550063324,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.5857,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.8733482336806271,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.5395,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 1.5754667694130446,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.8551,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.026619296346142,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.5609,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.9301919561276994,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.5755,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 0.8790969152130147,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.5269,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 0.8909937598519407,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.5121,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.6561997149156265,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.4948,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 0.6912117623767908,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.5176,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 0.7437610952904947,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.4914,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.7026861026072948,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.4788,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 0.5809296387083083,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.4816,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 1.0107791402531994,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.6794,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.6424539355378337,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.4825,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 0.7542031622647473,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.4446,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 0.6714440923614907,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.4194,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.5821287458120061,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.4674,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 0.6131543708192907,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.4467,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 0.6330509230082857,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.4316,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.6271528911438414,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.435,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 0.5302978616627516,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.3998,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 0.8739795045555382,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.684,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.5734268578269217,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.3893,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.5779200174086098,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.3813,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 0.5499093804843386,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.4055,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.5265167844202177,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.3812,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 0.5400291629686085,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.3753,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 0.5106433176681838,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.375,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.4566746058161919,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.3629,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.4632133044636238,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.382,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.526047149406831,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3743,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.8126085232325019,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.5711,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.4455087514591805,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.3194,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.5287858156135754,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.339,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.5187534527789833,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.3151,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.48548565527293497,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.3196,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.575974481147825,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.3338,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.5546881944998475,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.3117,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.4307213262189174,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.3083,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.44082398239007436,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.3231,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.8402823514466229,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.4957,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.4874816730227165,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.2668,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.5376664005455791,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.2668,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.5419620482445535,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.2833,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.5257415036659032,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.2675,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.5024353585089631,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.2761,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.48228509714201534,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.2665,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.562477591851496,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.2755,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.5102782233861666,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.2754,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.49093407572541736,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.2573,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 0.7479428740498759,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.3979,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 0.5201056231969392,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.22,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.5151144705091305,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.2389,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 0.49602585283187683,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.2457,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 0.6035258985510871,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.2281,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.506215143501012,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.2147,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.4965428798020858,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.2455,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.44190339502547793,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.2018,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.45547876258399317,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.1986,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 0.7161587795863175,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.3547,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.42601775992677765,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.2076,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.4944626930087208,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.1929,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.5728528367532345,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.2035,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.4362550178442035,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.1989,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.45868257290328096,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.1949,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.4637782410909022,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.1844,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.44510439455018846,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.1943,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.4461857643014094,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.1945,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.4082427904834379,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.1673,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 0.6395003552555746,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.2712,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.47692878154869267,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.1623,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.43643323518964133,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.1694,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.3903382248014779,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.1488,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 0.47851943237177375,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.1789,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.4548987582200087,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.1623,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.4425566552741696,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.1663,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.42661164836114646,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.1599,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.41370944162867496,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.1684,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 0.7267937888823227,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.2643,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 0.45995032284252424,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.1531,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.4406434429607988,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.1605,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.42714155253941855,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.1623,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.39220634705008045,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.1441,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.47555340870435764,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.1598,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.5228165541628115,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.1628,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.43073214177365693,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.1519,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.37897080903635194,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.1428,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.371960658859043,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.1406,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.6022907841646971,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.2183,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.38744645851536486,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.1563,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.3695125100205556,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.142,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.40210728163448434,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.1365,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.4001408002590834,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.1482,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.3655046893382892,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.1312,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.3771122612835669,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.1458,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.3478381459734836,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.1662,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.34843621011773535,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.1446,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.596966030465398,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.2275,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.3605724857145422,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.1392,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.36064053210086594,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.1442,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.3551070685344023,
      "learning_rate": 0.0,
      "loss": 0.1526,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 1.4430009490721997e+17,
      "train_loss": 0.3592390182436022,
      "train_runtime": 4910.5215,
      "train_samples_per_second": 0.837,
      "train_steps_per_second": 0.024
    }
  ],
  "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": 1.4430009490721997e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}