{
  "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": 6.853755327038079,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.054,
      "step": 1
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 6.7248331828080925,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0479,
      "step": 2
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 6.793346657766399,
      "learning_rate": 2.5e-06,
      "loss": 1.076,
      "step": 3
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 6.399417172855183,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0488,
      "step": 4
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 4.973638656499535,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.0113,
      "step": 5
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 3.1828149939077117,
      "learning_rate": 5e-06,
      "loss": 0.9835,
      "step": 6
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 2.6352548487400074,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9569,
      "step": 7
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 3.8080928407525545,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9363,
      "step": 8
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 4.275332292538442,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9485,
      "step": 9
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 6.873397157307676,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.5612,
      "step": 10
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 3.3656617520903636,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.7947,
      "step": 11
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 3.524663093579077,
      "learning_rate": 1e-05,
      "loss": 0.9017,
      "step": 12
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 2.5091723070752003,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.8621,
      "step": 13
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 1.6561631908665362,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.7153,
      "step": 14
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 1.7934912386240331,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.7946,
      "step": 15
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 2.1179200863384553,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.9096,
      "step": 16
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 1.4608060819959077,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.7832,
      "step": 17
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 1.2126476930576133,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.7584,
      "step": 18
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 2.146192940889719,
      "learning_rate": 9.890738003669029e-06,
      "loss": 1.1585,
      "step": 19
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 1.369292660595033,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.7294,
      "step": 20
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 1.2626340360076094,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.7097,
      "step": 21
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 0.9597691127181127,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.6938,
      "step": 22
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 1.2604895009103103,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.7032,
      "step": 23
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 1.2688720086624306,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.6648,
      "step": 24
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 1.0868419667652194,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.7288,
      "step": 25
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 1.3292378755288756,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.6912,
      "step": 26
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 1.2346104887882967,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.6945,
      "step": 27
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 1.02561670902633,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.6629,
      "step": 28
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 1.9356146978360425,
      "learning_rate": 9.36704100308565e-06,
      "loss": 1.005,
      "step": 29
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 1.271249473510972,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.6001,
      "step": 30
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 1.0253844700862984,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.5946,
      "step": 31
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 0.9114775970910892,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.566,
      "step": 32
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 1.22530795109878,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.7532,
      "step": 33
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 0.6946604067078974,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.4619,
      "step": 34
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 1.047406831639789,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.6378,
      "step": 35
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.711823766609595,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.5243,
      "step": 36
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 0.9122786016111567,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.6424,
      "step": 37
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 1.3017517797000289,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.8755,
      "step": 38
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.8336965053721384,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.5398,
      "step": 39
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.7646224552052723,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.5382,
      "step": 40
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 0.7875281400060778,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.5356,
      "step": 41
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.845564019017656,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.5194,
      "step": 42
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 0.7433701271634005,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.5147,
      "step": 43
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 0.9200656896721353,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.4964,
      "step": 44
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.8863920970016912,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.483,
      "step": 45
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.805646965467507,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.4888,
      "step": 46
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.8469077839570921,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.473,
      "step": 47
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 1.1316845580084651,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.7226,
      "step": 48
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.7610540567385804,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.4786,
      "step": 49
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.6395383235541354,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.3273,
      "step": 50
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.8554027282099782,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.4886,
      "step": 51
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.8451926273461702,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.4818,
      "step": 52
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.6633671884643004,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.3512,
      "step": 53
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.7877500797629732,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.3786,
      "step": 54
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.7686923486972509,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.4644,
      "step": 55
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.65164216458978,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.3689,
      "step": 56
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 1.3973623392345897,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.6268,
      "step": 57
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.8029150370176054,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.386,
      "step": 58
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.8520885652834594,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.3201,
      "step": 59
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 1.024212957711499,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.3481,
      "step": 60
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.9772378939893376,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.3466,
      "step": 61
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.8194978346963566,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.3546,
      "step": 62
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.8534533201642622,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.314,
      "step": 63
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.7809321319644733,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.3236,
      "step": 64
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.9096044136211295,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.3308,
      "step": 65
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.8958045974543299,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.3232,
      "step": 66
    },
    {
      "epoch": 7.050632911392405,
      "grad_norm": 1.2284199200433397,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.455,
      "step": 67
    },
    {
      "epoch": 7.151898734177215,
      "grad_norm": 1.0209394478690026,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.3172,
      "step": 68
    },
    {
      "epoch": 7.253164556962025,
      "grad_norm": 0.8411127195413696,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.2454,
      "step": 69
    },
    {
      "epoch": 7.3544303797468356,
      "grad_norm": 1.007359276104836,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.2928,
      "step": 70
    },
    {
      "epoch": 7.455696202531645,
      "grad_norm": 1.3583271315700094,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.2754,
      "step": 71
    },
    {
      "epoch": 7.556962025316456,
      "grad_norm": 0.8444737453345904,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.2774,
      "step": 72
    },
    {
      "epoch": 7.658227848101266,
      "grad_norm": 0.7995027858063242,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.2268,
      "step": 73
    },
    {
      "epoch": 7.759493670886076,
      "grad_norm": 0.829822765554238,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.2376,
      "step": 74
    },
    {
      "epoch": 7.860759493670886,
      "grad_norm": 0.7949993356800901,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.2417,
      "step": 75
    },
    {
      "epoch": 7.962025316455696,
      "grad_norm": 1.083008382706899,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.3908,
      "step": 76
    },
    {
      "epoch": 8.10126582278481,
      "grad_norm": 0.7384347404687762,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.2232,
      "step": 77
    },
    {
      "epoch": 8.20253164556962,
      "grad_norm": 0.7489172508921755,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.2237,
      "step": 78
    },
    {
      "epoch": 8.30379746835443,
      "grad_norm": 0.7997609752723011,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.2137,
      "step": 79
    },
    {
      "epoch": 8.405063291139241,
      "grad_norm": 0.7934569206656524,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.2025,
      "step": 80
    },
    {
      "epoch": 8.50632911392405,
      "grad_norm": 0.7246422597431458,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.1998,
      "step": 81
    },
    {
      "epoch": 8.60759493670886,
      "grad_norm": 0.6754392616504892,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.2075,
      "step": 82
    },
    {
      "epoch": 8.708860759493671,
      "grad_norm": 0.6981132485392166,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.2269,
      "step": 83
    },
    {
      "epoch": 8.810126582278482,
      "grad_norm": 0.7631604441829668,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.2012,
      "step": 84
    },
    {
      "epoch": 8.91139240506329,
      "grad_norm": 0.691094769682961,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.2054,
      "step": 85
    },
    {
      "epoch": 9.050632911392405,
      "grad_norm": 1.044439601871888,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.3057,
      "step": 86
    },
    {
      "epoch": 9.151898734177216,
      "grad_norm": 0.7389192407768973,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.1753,
      "step": 87
    },
    {
      "epoch": 9.253164556962025,
      "grad_norm": 0.8928754224532668,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.1839,
      "step": 88
    },
    {
      "epoch": 9.354430379746836,
      "grad_norm": 0.7174544029609661,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.1495,
      "step": 89
    },
    {
      "epoch": 9.455696202531646,
      "grad_norm": 1.0095542055134636,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.1764,
      "step": 90
    },
    {
      "epoch": 9.556962025316455,
      "grad_norm": 0.9898931188613699,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.1849,
      "step": 91
    },
    {
      "epoch": 9.658227848101266,
      "grad_norm": 0.6806978513304183,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.1757,
      "step": 92
    },
    {
      "epoch": 9.759493670886076,
      "grad_norm": 0.7369725238345852,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.1686,
      "step": 93
    },
    {
      "epoch": 9.860759493670885,
      "grad_norm": 0.8401916486657559,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.1922,
      "step": 94
    },
    {
      "epoch": 9.962025316455696,
      "grad_norm": 1.1445038380201404,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.2362,
      "step": 95
    },
    {
      "epoch": 10.10126582278481,
      "grad_norm": 1.713991282049752,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.178,
      "step": 96
    },
    {
      "epoch": 10.20253164556962,
      "grad_norm": 0.7263920249731424,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.1472,
      "step": 97
    },
    {
      "epoch": 10.30379746835443,
      "grad_norm": 0.654115196746866,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.1498,
      "step": 98
    },
    {
      "epoch": 10.405063291139241,
      "grad_norm": 0.6984926674484835,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.16,
      "step": 99
    },
    {
      "epoch": 10.50632911392405,
      "grad_norm": 0.779812844768717,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.1501,
      "step": 100
    },
    {
      "epoch": 10.60759493670886,
      "grad_norm": 0.8903247160522805,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.1474,
      "step": 101
    },
    {
      "epoch": 10.708860759493671,
      "grad_norm": 0.8749237869088703,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.1361,
      "step": 102
    },
    {
      "epoch": 10.810126582278482,
      "grad_norm": 0.9183489788922128,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.1635,
      "step": 103
    },
    {
      "epoch": 10.91139240506329,
      "grad_norm": 0.640927724740078,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.1611,
      "step": 104
    },
    {
      "epoch": 11.050632911392405,
      "grad_norm": 0.9234957345398533,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.2163,
      "step": 105
    },
    {
      "epoch": 11.151898734177216,
      "grad_norm": 0.5850183080908703,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.1532,
      "step": 106
    },
    {
      "epoch": 11.253164556962025,
      "grad_norm": 0.623097599780791,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.1417,
      "step": 107
    },
    {
      "epoch": 11.354430379746836,
      "grad_norm": 0.8886870572165986,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.1452,
      "step": 108
    },
    {
      "epoch": 11.455696202531646,
      "grad_norm": 0.6809910647284587,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.1502,
      "step": 109
    },
    {
      "epoch": 11.556962025316455,
      "grad_norm": 0.6186409670002029,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.1259,
      "step": 110
    },
    {
      "epoch": 11.658227848101266,
      "grad_norm": 0.6414798273261219,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.1307,
      "step": 111
    },
    {
      "epoch": 11.759493670886076,
      "grad_norm": 0.5732446786349397,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.1354,
      "step": 112
    },
    {
      "epoch": 11.860759493670885,
      "grad_norm": 0.5596747757561155,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.1331,
      "step": 113
    },
    {
      "epoch": 11.962025316455696,
      "grad_norm": 0.8922015800275596,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.2331,
      "step": 114
    },
    {
      "epoch": 12.10126582278481,
      "grad_norm": 0.5784867209037564,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.1322,
      "step": 115
    },
    {
      "epoch": 12.20253164556962,
      "grad_norm": 0.5683118055283999,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.1465,
      "step": 116
    },
    {
      "epoch": 12.30379746835443,
      "grad_norm": 0.5300130325739068,
      "learning_rate": 0.0,
      "loss": 0.1496,
      "step": 117
    },
    {
      "epoch": 12.30379746835443,
      "step": 117,
      "total_flos": 9.101952189202432e+16,
      "train_loss": 0.4595338146910708,
      "train_runtime": 4622.163,
      "train_samples_per_second": 0.889,
      "train_steps_per_second": 0.025
    }
  ],
  "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": 9.101952189202432e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}