{
  "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": 8.263133428716758,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.2648,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 8.227005916593942,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.2436,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 7.803593120207985,
      "learning_rate": 2.5e-06,
      "loss": 1.1862,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 7.849258258663947,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.2503,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 6.2753083586492835,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.2023,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 3.4263617166084503,
      "learning_rate": 5e-06,
      "loss": 1.0913,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 3.261457774980699,
      "learning_rate": 5.833333333333334e-06,
      "loss": 1.1178,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 4.142984185527025,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.0709,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 4.367184321709739,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.0325,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 4.259432671021632,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.0254,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 3.9944873006407877,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.9362,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 3.9143947005709188,
      "learning_rate": 1e-05,
      "loss": 0.9557,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 3.3802693510864374,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.9489,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 2.704648184735901,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.8758,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 2.50049571017678,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.8645,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 2.967736412557229,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.8685,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 2.3765379256472534,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.8832,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 2.0143565095320453,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.8676,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 2.0622470855912813,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.7951,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.7272387088018006,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.7057,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 1.686993472949459,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.6444,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 1.7266502763749563,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.6774,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 1.4830294094959673,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.6296,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 1.4024669784427999,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.6691,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 1.3484466112925524,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.6493,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 1.3701196775492421,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.6349,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 1.2084486250365292,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.6451,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 1.2382599604468796,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.5924,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 1.2832705547316599,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.5783,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 1.2273471382031556,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.4682,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 1.031611070746068,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.4594,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 1.3120483583350326,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.4061,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 1.5691842908478648,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.4126,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 1.1437098636798506,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.4428,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 1.1261599416589012,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.4399,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 1.2245833075064751,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.4017,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 1.140227654907964,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.4211,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 1.1060858687366057,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.4628,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.9834293980598223,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.3399,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.9910627675055094,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.3337,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 1.0192068434141106,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.2346,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 1.2951533901697594,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.242,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 1.4511216070625947,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.3284,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 1.0256947326462187,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.2473,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 1.04989322520992,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.3013,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 1.0193597997894372,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.28,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 1.0940908131948774,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.2624,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.9328197055502845,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.238,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.9752436705868058,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.2106,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.8646794067651836,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.1731,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.9587648137845494,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.1796,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 1.2638039964827255,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.1789,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 1.4068957372169308,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.1901,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.9508506078517881,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.1714,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.8982646159065789,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.1539,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.9461348938153032,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.2085,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.8586250409348295,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.1893,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.9228772291908832,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.1767,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.7447999643177767,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.1415,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.6043984301773121,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.1266,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.7469933492225036,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.1683,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.7667764527692278,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.0792,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.8900050256776663,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.0893,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.7816433028923908,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.0983,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.9343519855906365,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.1191,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.6914501564830962,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.1027,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 0.6575536282897345,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.103,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 0.5566391545035563,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.0803,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.6993259031792627,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.098,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 0.6306064669096859,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.1188,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 0.554445687131775,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.0939,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.6072015980872499,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.0909,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.6123687095622097,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.0947,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.6250157041616546,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.0671,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.6516027711246847,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.0818,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 0.5520273117193212,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.0423,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.4280903647623177,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.0737,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.534720074097009,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.077,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.40358359011350287,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.0502,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.5142404001843289,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.0784,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.36267975715537165,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.056,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.3952878367366781,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.0744,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.36916874795080523,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.0321,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.28175811293713127,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.0165,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.4973943162833611,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.0665,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 0.397230346711515,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.0792,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.34148928551541297,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.0529,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.29345535413421,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.0404,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.3062297780923687,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.0377,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 0.3231187906029722,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.0375,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.38110736018487873,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.0784,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.33016280620963717,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.0552,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.4422130347129991,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.0384,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.30090735160057325,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.0397,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 0.36921917303858576,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.0607,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 0.3247019344660354,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.0436,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.2240500263996633,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.0415,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.21438884589476023,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.0227,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.2865798449203878,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.0442,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.2399517159094271,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.0493,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.2984334754248065,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.0446,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.22587484176009187,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.0323,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.3460992395024099,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.0649,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.3441447458490397,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.0371,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.2943181630661656,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.0377,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.17540738811189438,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.0182,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.1956389336635979,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.0141,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.2633122843936609,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.0312,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.2694380916528542,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.0543,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.34064345414471936,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.0755,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.24603717218515656,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.0346,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.2644942780388346,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.038,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.2505348195813056,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.047,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.29824708107593695,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.0471,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.24197332725743803,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.0446,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.2684036452944054,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.0567,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.30424403266769756,
      "learning_rate": 0.0,
      "loss": 0.0453,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 3.187488584420557e+16,
      "train_loss": 0.32672480929197156,
      "train_runtime": 1915.8867,
      "train_samples_per_second": 2.144,
      "train_steps_per_second": 0.061
    }
  ],
  "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": 3.187488584420557e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}