{
  "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": 6.609177358582297,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.0897,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 6.649744350714212,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0727,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 6.301279727095918,
      "learning_rate": 2.5e-06,
      "loss": 0.9883,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 6.258119588505225,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0626,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 5.470179860969025,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.0482,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 3.283015838939176,
      "learning_rate": 5e-06,
      "loss": 1.0581,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 2.1947638321789094,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9446,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 4.081439176202081,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9181,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 4.410147130916803,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9532,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 4.540776464395203,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9346,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 3.885615542747028,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8288,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 4.053974245643597,
      "learning_rate": 1e-05,
      "loss": 0.8851,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 3.025464243351496,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.8292,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 2.736088250738933,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.8384,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 2.668966980209538,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.8068,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 2.143206456393247,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.7245,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 2.4549083973719643,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.8384,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 2.0804418098318713,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.774,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 1.9415014052867388,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.7983,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.597846369746843,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.6952,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 1.6484213709018318,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.6584,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 1.689507554025834,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.6775,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 1.277915953096299,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.6032,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 1.1092820058361665,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.6447,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 1.2562418830755544,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.6088,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 1.4557131494951154,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.606,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 1.1861706605228486,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.6391,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 0.9904739133949982,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.611,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 1.2696893899862918,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.6091,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 1.0902562261761413,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.5194,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 0.911800422526914,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.4847,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 1.1187653106000894,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.5099,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 1.1770489836169016,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.5065,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 0.8741723159857043,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.4707,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 0.9915349782791263,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.4793,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.9593823515716747,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.4477,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 0.9045272272312278,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.4365,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 1.182033351912803,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.4648,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.8182323794145082,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.3927,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.8905072957103408,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.3374,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 0.819415788173681,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.3799,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.8219617365410626,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.3661,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 1.129632782336193,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.3204,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 0.9472562459234598,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.3028,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 1.0145116552050686,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.3154,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.9001061690961518,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.3423,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.9011158451360376,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3813,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.8205130050955913,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.3418,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.674871028894581,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.2778,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.833919604837176,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.249,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.7678280921240637,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.2854,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.710112407575669,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.2856,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.9347755222448465,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.2512,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.665853209779997,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.27,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.8452891304804028,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.245,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.8649059943815952,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.222,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.9166123499965918,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.266,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.7920120511801666,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.2038,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.770631643136936,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.2193,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.5831733051789565,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.22,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.8527277857149947,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.2117,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.7704613880100631,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.223,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.7773907175981413,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.174,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.7769749189896199,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.1868,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.8267884335813923,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.2136,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.7928454596431526,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.1389,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 0.7709556289100814,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.1966,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 0.5510692381534656,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.1804,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.6278460752331894,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.1647,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 0.6100165944754435,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.1094,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 0.6463198803930352,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.1782,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.6448361149000619,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.1277,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.6040303590541892,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.1751,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.5587852678819757,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.1542,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.6413497450213962,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.1521,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 0.5537578504838808,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.1161,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.5601665115628431,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.1396,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.5270129442086667,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.1407,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.4201420309517452,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.0915,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.5268326582855813,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.1215,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.5707391958242994,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.1391,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.5730303145069258,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.122,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.570187470897077,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.1109,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.5074740455258397,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.1093,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.5638397468629261,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.0818,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 0.4926715072002825,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.1355,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.6850064120842905,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.1414,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.591284436336564,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.1054,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.42105669663994355,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.1118,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 0.4208572972620201,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.0979,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.5811612562775901,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.1214,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.5611914929736412,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.1313,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.4749176595038225,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.0798,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.49489973199370213,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.102,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 0.46488381887894914,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.0628,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 0.3758483445602909,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.0608,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.31430205805431016,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.0613,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.3652078608129394,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.0999,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.37784728021110336,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.093,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.3975802358034184,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.0883,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.40557780163105983,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.0928,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.3756746340932111,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.1019,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.43993322646417904,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.1388,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.3550643375382574,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.1282,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.3818699688120338,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.0633,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.2929804644439644,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.0612,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.3674622882796918,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.0956,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.3580272530776606,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.0962,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.32263573306650506,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.0993,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.3287241802551749,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.0759,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.33557013513750295,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.0778,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.38621705302714415,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.1233,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.317690622375799,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.0746,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.3169009479721045,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.0936,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.3542145906248746,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.1221,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.3243879231853871,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.0985,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.3834052287774511,
      "learning_rate": 0.0,
      "loss": 0.087,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 6.247445538904474e+16,
      "train_loss": 0.35404415571918857,
      "train_runtime": 5493.3154,
      "train_samples_per_second": 0.748,
      "train_steps_per_second": 0.021
    }
  ],
  "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": 6.247445538904474e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}