{
  "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": 7.680033192188056,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.876,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 7.605076623290318,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.8658,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 6.778648218751108,
      "learning_rate": 2.5e-06,
      "loss": 0.8177,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 6.509762542649743,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.7961,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 5.60547296324972,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.8073,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 2.9603418690542127,
      "learning_rate": 5e-06,
      "loss": 0.7123,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 2.6586350137017654,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.719,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 3.628503370339582,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.6693,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 3.994437575777995,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6902,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 3.8311013222873855,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.6371,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 4.300806163841766,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.6389,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 3.763694085806352,
      "learning_rate": 1e-05,
      "loss": 0.5804,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 2.9051031343691744,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.5677,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 1.9919406509125026,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.5447,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 2.214405561229308,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.5839,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 2.4360157076781483,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.5685,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 2.266012553917941,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.5271,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 1.6843183104054464,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.4994,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 1.088950454916664,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.5071,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.200233336627422,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.4746,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 1.2637691080193945,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.4618,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 1.3769873879592567,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.4508,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 0.8682440711586037,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.3964,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.9509335750395728,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.4052,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 1.1726269272210401,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.4393,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 1.0363716677545973,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.4439,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.8229007452516713,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.3915,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 0.9246655473068639,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.391,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 0.8702250576114761,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.4052,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.8951505231845288,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.3665,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 0.8202167130655299,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.334,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 0.7812315273644796,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.321,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.9882162553157401,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.2901,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 0.8109137436623326,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.2626,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 0.7499397799268233,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.3374,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.8565071413317179,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.2619,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 0.7524729309779767,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.2689,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 0.7531983387104568,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.3241,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.6868113396123745,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.2336,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.6794221220377648,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.2686,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 0.7127200441976037,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.2302,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.9320401873680159,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.2307,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 0.741263544835177,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.2473,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 0.6812338139520304,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.1837,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.7924153987851118,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.2433,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.704653416654649,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.226,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.7727681689903335,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.265,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.6187802591301362,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.1834,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.5876280442306084,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.1881,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.672959788742261,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.1319,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.711584893214224,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.1769,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.7178272865398647,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.2074,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.7830848244454021,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.2037,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.6179328370518615,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.1574,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.6670115628975251,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.161,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.6785055355731117,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.1791,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.6338228496221349,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.1689,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.5540510682761942,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.1528,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.5818482838873013,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.1912,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.6163781450110954,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.1284,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.677574989780631,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.1589,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.6435283762104541,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.1194,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.6239949273645631,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.1045,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.572687410995689,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.1636,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.5449061137184503,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.162,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.5475880191654193,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.0917,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 0.6385594435677409,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.0986,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 0.5885768339308608,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.1327,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.46414251730482564,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.117,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 0.4596678374134551,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.1052,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 0.6382991489091762,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.1396,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.5234765183562552,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.0927,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.5028097125423595,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.1039,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.5612361364934108,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.1089,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.5921818102172546,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.0905,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 0.7756744852324812,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.1256,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.46067091094544693,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.0695,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.5331891345046547,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.1007,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.38056227705123474,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.1152,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.47749834030257926,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.1434,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.33181359449202125,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.0601,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.3692812970351156,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.0461,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.5036817238632461,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.0994,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.5549186139256994,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.0947,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.5312869981574345,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.0991,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 0.3759462356183745,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.1179,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.33383214559138846,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.0762,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.26312712916539965,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.0745,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.30730572593714295,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.0754,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 0.31093551470161,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.0825,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.3115908963835014,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.0856,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.3345152880560803,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.0652,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.32769619009343287,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.0816,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.4334283532490735,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.0869,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 0.2877036407917396,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.077,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 0.24167386699272236,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.0721,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.19556579405259916,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.0383,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.2552062802729188,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.0824,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.2898550507651794,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.0874,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.4524836615561104,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.0519,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.2859763217293582,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.1038,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.2740176489434147,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.0724,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.28031594418514205,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.0913,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.29825597093016865,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.0667,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.26440161804086987,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.083,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.2264550862330323,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.0452,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.20794688488652271,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.0375,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.28341702236760624,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.1221,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.2432505235831095,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.0469,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.23671574535859555,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.0635,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.2882576880295542,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.1207,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.21717624854035625,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.073,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.25992837878278163,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.0677,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.27299775911291413,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.0789,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.22631594857909607,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.0663,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.3003461315969225,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.0834,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.21311917276522419,
      "learning_rate": 0.0,
      "loss": 0.0534,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 5.484038039850189e+16,
      "train_loss": 0.2492905456540931,
      "train_runtime": 3417.7179,
      "train_samples_per_second": 1.202,
      "train_steps_per_second": 0.034
    }
  ],
  "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": 5.484038039850189e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}