{
  "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": 5.895437431587112,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.8025,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 6.55011655692263,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.8798,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 6.328385102296379,
      "learning_rate": 2.5e-06,
      "loss": 0.8668,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 5.617604720367703,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.8131,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 4.572290590133521,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.7931,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 2.817935556711524,
      "learning_rate": 5e-06,
      "loss": 0.8143,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 2.343065328802187,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.7461,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 4.030076959055218,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.7837,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 4.116892800507118,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7347,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 6.010861822444835,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.0788,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 4.193504321170482,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.7081,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 4.348430450002865,
      "learning_rate": 1e-05,
      "loss": 0.7532,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 2.5333157696365998,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.6739,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 1.8353886386232434,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.6103,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 2.372649173717505,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.6793,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 2.1879699813546067,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.5917,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 1.2387967220243845,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.5119,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 1.3155704193489053,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.7505,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 1.9861419098496649,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.8391,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.2872463511211474,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.5918,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 1.0115181200061254,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.6163,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 0.9102784864819136,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.537,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 0.7723417034750111,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.5367,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.8783764489730859,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.5329,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 0.8782262047190333,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.5452,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 0.7389522197168812,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.4942,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.8121085712863032,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.4927,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 0.7618741438884619,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.5312,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 1.2718116583571646,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.7125,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.6500980617895746,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.4407,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 0.8261449456501613,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.5216,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 0.7149726839502812,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.4257,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.7287392642072873,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.4487,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 0.6926625472743289,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.4269,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 0.7430355525518108,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.4594,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.6951675618053621,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.4411,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 0.8131115512274725,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.4851,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 0.9746912551804889,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.6522,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.6135819521316205,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.3605,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.771617070195418,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.3884,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 0.7403480974516429,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.3899,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.6487685272560921,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.3628,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 0.6814856632845298,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.369,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 0.7543767184996715,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.3729,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.6844089474834384,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.3336,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.7901863608944344,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.371,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.6753366404746309,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3695,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.9374564193170443,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.4896,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.6734302320117497,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.3232,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.5853443311927671,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.2417,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.8431653133600792,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.3806,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.6063521346545561,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.2492,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.6633606297212277,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.3119,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.6688638883108692,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.2968,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.6424424076730402,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.2951,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.5924677605310528,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.2551,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.9869794559950298,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.4196,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.6499484367851245,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.2624,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.6415770304364082,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.2329,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.8307170104602447,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.2311,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.7447130507587343,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.2186,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.5778382545568016,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.2146,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.7052886016028709,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.2163,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.7154994863489477,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.2283,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.643992126658529,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.221,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.6724133134564504,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.2085,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 1.160211427627962,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.3412,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 0.6780483064743578,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.1891,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.780535064628723,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.1885,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 0.5480518449729264,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.1405,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 0.8075066487029843,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.1995,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.956446937600229,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.1708,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.510378350157666,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.131,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.656821227766225,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.1576,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.7136751454302225,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.1818,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 1.1209877010915406,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.2423,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.6899721837091559,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.1412,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.5290128885245698,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.1172,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.9518167488224024,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.1349,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.7551500813610078,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.1489,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.5985052847049391,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.1203,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.5751197656242599,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.1245,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.6383274566722233,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.1496,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.5723593804098357,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.1352,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.6677978416090448,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.1453,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 0.8960387547513665,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.1754,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.6722939232433549,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.1084,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.4394195937877518,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.0788,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.5305089443300481,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.1111,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 0.6465935413812988,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.1283,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.6819902452141822,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.0843,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.6373388779124766,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.1159,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.5123439890548894,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.0894,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.5860372625400543,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.1411,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 0.7835229137826922,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.1371,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 0.635322382492641,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.0946,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.5697842838240275,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.0856,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.5900259121677638,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.1062,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.5655488810994538,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.1031,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.44428290243971363,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.0913,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.4300060717644238,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.0987,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.48971617650201743,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.0811,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.4687521218771687,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.0858,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.5160899848695946,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.0838,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.9201523755739287,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.1569,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.4215647228430539,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.0927,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.4270383042089781,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.0738,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.4144363930867899,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.0712,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.3748170921055616,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.0759,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.4568040505231358,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.0889,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.3600441942152769,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.0651,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.4451833224686696,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.0981,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.42144624506197254,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.0997,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.8927133255766454,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.1527,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.38943910408029747,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.0872,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.40225895597156897,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.0769,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.3875069928114505,
      "learning_rate": 0.0,
      "loss": 0.1017,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 8.19186747005993e+16,
      "train_loss": 0.33621527120853084,
      "train_runtime": 3901.6974,
      "train_samples_per_second": 1.053,
      "train_steps_per_second": 0.03
    }
  ],
  "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": 8.19186747005993e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}