{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9872340425531916,
  "eval_steps": 500,
  "global_step": 234,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01276595744680851,
      "grad_norm": 6.353932376962155,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 1.1912,
      "step": 1
    },
    {
      "epoch": 0.02553191489361702,
      "grad_norm": 6.22538335218383,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.1524,
      "step": 2
    },
    {
      "epoch": 0.03829787234042553,
      "grad_norm": 6.023789244570738,
      "learning_rate": 1.25e-06,
      "loss": 1.1156,
      "step": 3
    },
    {
      "epoch": 0.05106382978723404,
      "grad_norm": 6.298814184785475,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.2029,
      "step": 4
    },
    {
      "epoch": 0.06382978723404255,
      "grad_norm": 5.954891137312739,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 1.1737,
      "step": 5
    },
    {
      "epoch": 0.07659574468085106,
      "grad_norm": 4.707769789262157,
      "learning_rate": 2.5e-06,
      "loss": 1.1566,
      "step": 6
    },
    {
      "epoch": 0.08936170212765958,
      "grad_norm": 4.06547531235841,
      "learning_rate": 2.916666666666667e-06,
      "loss": 1.0835,
      "step": 7
    },
    {
      "epoch": 0.10212765957446808,
      "grad_norm": 2.6783100585231767,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0894,
      "step": 8
    },
    {
      "epoch": 0.1148936170212766,
      "grad_norm": 2.531714805749627,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 1.1027,
      "step": 9
    },
    {
      "epoch": 0.1276595744680851,
      "grad_norm": 2.3328797912721426,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.0661,
      "step": 10
    },
    {
      "epoch": 0.14042553191489363,
      "grad_norm": 4.06134025571158,
      "learning_rate": 4.583333333333333e-06,
      "loss": 1.0376,
      "step": 11
    },
    {
      "epoch": 0.15319148936170213,
      "grad_norm": 4.184622856911514,
      "learning_rate": 5e-06,
      "loss": 1.0636,
      "step": 12
    },
    {
      "epoch": 0.16595744680851063,
      "grad_norm": 4.199138698508921,
      "learning_rate": 5.416666666666667e-06,
      "loss": 1.0713,
      "step": 13
    },
    {
      "epoch": 0.17872340425531916,
      "grad_norm": 3.0771186783481066,
      "learning_rate": 5.833333333333334e-06,
      "loss": 1.021,
      "step": 14
    },
    {
      "epoch": 0.19148936170212766,
      "grad_norm": 2.9827797840666106,
      "learning_rate": 6.25e-06,
      "loss": 0.994,
      "step": 15
    },
    {
      "epoch": 0.20425531914893616,
      "grad_norm": 2.5267499262248165,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9823,
      "step": 16
    },
    {
      "epoch": 0.2170212765957447,
      "grad_norm": 2.2630049185041625,
      "learning_rate": 7.083333333333335e-06,
      "loss": 0.9355,
      "step": 17
    },
    {
      "epoch": 0.2297872340425532,
      "grad_norm": 2.1526932427229086,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9551,
      "step": 18
    },
    {
      "epoch": 0.2425531914893617,
      "grad_norm": 1.8518564917059142,
      "learning_rate": 7.916666666666667e-06,
      "loss": 0.9394,
      "step": 19
    },
    {
      "epoch": 0.2553191489361702,
      "grad_norm": 1.6748374760293612,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9586,
      "step": 20
    },
    {
      "epoch": 0.2680851063829787,
      "grad_norm": 1.4038164790827745,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.9333,
      "step": 21
    },
    {
      "epoch": 0.28085106382978725,
      "grad_norm": 1.3518569082484015,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.9325,
      "step": 22
    },
    {
      "epoch": 0.2936170212765957,
      "grad_norm": 1.1090409192234645,
      "learning_rate": 9.583333333333335e-06,
      "loss": 0.8934,
      "step": 23
    },
    {
      "epoch": 0.30638297872340425,
      "grad_norm": 1.1504957844523982,
      "learning_rate": 1e-05,
      "loss": 0.9226,
      "step": 24
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 1.0220624504910976,
      "learning_rate": 9.999440509051367e-06,
      "loss": 0.8972,
      "step": 25
    },
    {
      "epoch": 0.33191489361702126,
      "grad_norm": 1.0571307771486256,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.8894,
      "step": 26
    },
    {
      "epoch": 0.3446808510638298,
      "grad_norm": 0.9979834526620437,
      "learning_rate": 9.994965332706574e-06,
      "loss": 0.8751,
      "step": 27
    },
    {
      "epoch": 0.3574468085106383,
      "grad_norm": 0.9631038384731712,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.8794,
      "step": 28
    },
    {
      "epoch": 0.3702127659574468,
      "grad_norm": 0.9639536468707164,
      "learning_rate": 9.986018985905901e-06,
      "loss": 0.8651,
      "step": 29
    },
    {
      "epoch": 0.3829787234042553,
      "grad_norm": 0.9991789301222855,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.8748,
      "step": 30
    },
    {
      "epoch": 0.39574468085106385,
      "grad_norm": 0.7084177291168907,
      "learning_rate": 9.972609476841368e-06,
      "loss": 0.8668,
      "step": 31
    },
    {
      "epoch": 0.4085106382978723,
      "grad_norm": 0.9941892229146153,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.9048,
      "step": 32
    },
    {
      "epoch": 0.42127659574468085,
      "grad_norm": 0.8881707523663802,
      "learning_rate": 9.954748808839675e-06,
      "loss": 0.8814,
      "step": 33
    },
    {
      "epoch": 0.4340425531914894,
      "grad_norm": 0.8130437377464066,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.8748,
      "step": 34
    },
    {
      "epoch": 0.44680851063829785,
      "grad_norm": 0.8570977832402222,
      "learning_rate": 9.932452969617607e-06,
      "loss": 0.8866,
      "step": 35
    },
    {
      "epoch": 0.4595744680851064,
      "grad_norm": 0.7423707329128824,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.8582,
      "step": 36
    },
    {
      "epoch": 0.4723404255319149,
      "grad_norm": 0.7198235435636622,
      "learning_rate": 9.905741916970863e-06,
      "loss": 0.8598,
      "step": 37
    },
    {
      "epoch": 0.4851063829787234,
      "grad_norm": 0.7749185329627992,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.8458,
      "step": 38
    },
    {
      "epoch": 0.4978723404255319,
      "grad_norm": 0.6433008471066703,
      "learning_rate": 9.874639560909118e-06,
      "loss": 0.8221,
      "step": 39
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 0.7978894394809969,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.8824,
      "step": 40
    },
    {
      "epoch": 0.5234042553191489,
      "grad_norm": 0.767845866948557,
      "learning_rate": 9.839173742253334e-06,
      "loss": 0.8744,
      "step": 41
    },
    {
      "epoch": 0.5361702127659574,
      "grad_norm": 0.7396968322828916,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.8635,
      "step": 42
    },
    {
      "epoch": 0.548936170212766,
      "grad_norm": 0.6779405395661932,
      "learning_rate": 9.799376207714446e-06,
      "loss": 0.8284,
      "step": 43
    },
    {
      "epoch": 0.5617021276595745,
      "grad_norm": 0.7153502005873587,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.8818,
      "step": 44
    },
    {
      "epoch": 0.574468085106383,
      "grad_norm": 0.6863317883746143,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.8713,
      "step": 45
    },
    {
      "epoch": 0.5872340425531914,
      "grad_norm": 0.7130901083688288,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.8134,
      "step": 46
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6898306425311603,
      "learning_rate": 9.706932333304518e-06,
      "loss": 0.8581,
      "step": 47
    },
    {
      "epoch": 0.6127659574468085,
      "grad_norm": 0.6982172738750606,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.8413,
      "step": 48
    },
    {
      "epoch": 0.625531914893617,
      "grad_norm": 0.8275251593813877,
      "learning_rate": 9.654368743221022e-06,
      "loss": 0.8602,
      "step": 49
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 0.7343850642149492,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.826,
      "step": 50
    },
    {
      "epoch": 0.6510638297872341,
      "grad_norm": 0.7654533248805878,
      "learning_rate": 9.597638862757255e-06,
      "loss": 0.8888,
      "step": 51
    },
    {
      "epoch": 0.6638297872340425,
      "grad_norm": 0.6500639945476345,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.8201,
      "step": 52
    },
    {
      "epoch": 0.676595744680851,
      "grad_norm": 0.87098438448592,
      "learning_rate": 9.536793472839325e-06,
      "loss": 0.8661,
      "step": 53
    },
    {
      "epoch": 0.6893617021276596,
      "grad_norm": 0.7053185152695439,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.807,
      "step": 54
    },
    {
      "epoch": 0.7021276595744681,
      "grad_norm": 0.672193427899284,
      "learning_rate": 9.471887038331686e-06,
      "loss": 0.8168,
      "step": 55
    },
    {
      "epoch": 0.7148936170212766,
      "grad_norm": 0.6999423156768017,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.8412,
      "step": 56
    },
    {
      "epoch": 0.7276595744680852,
      "grad_norm": 0.6568075220707117,
      "learning_rate": 9.40297765928369e-06,
      "loss": 0.8468,
      "step": 57
    },
    {
      "epoch": 0.7404255319148936,
      "grad_norm": 0.6590937008148525,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.8368,
      "step": 58
    },
    {
      "epoch": 0.7531914893617021,
      "grad_norm": 0.6343228191926094,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.8368,
      "step": 59
    },
    {
      "epoch": 0.7659574468085106,
      "grad_norm": 0.6711624025837197,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.8024,
      "step": 60
    },
    {
      "epoch": 0.7787234042553192,
      "grad_norm": 0.8360001836286379,
      "learning_rate": 9.253400328436699e-06,
      "loss": 0.8484,
      "step": 61
    },
    {
      "epoch": 0.7914893617021277,
      "grad_norm": 0.6685752235146978,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.8408,
      "step": 62
    },
    {
      "epoch": 0.8042553191489362,
      "grad_norm": 0.658221678510916,
      "learning_rate": 9.172866268606514e-06,
      "loss": 0.8535,
      "step": 63
    },
    {
      "epoch": 0.8170212765957446,
      "grad_norm": 0.6507254367865235,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.8747,
      "step": 64
    },
    {
      "epoch": 0.8297872340425532,
      "grad_norm": 0.6460028531302459,
      "learning_rate": 9.088596928322158e-06,
      "loss": 0.8235,
      "step": 65
    },
    {
      "epoch": 0.8425531914893617,
      "grad_norm": 0.6822964607405055,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.8686,
      "step": 66
    },
    {
      "epoch": 0.8553191489361702,
      "grad_norm": 0.6076116804305184,
      "learning_rate": 9.000667740056033e-06,
      "loss": 0.8193,
      "step": 67
    },
    {
      "epoch": 0.8680851063829788,
      "grad_norm": 0.6590919977095543,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.8728,
      "step": 68
    },
    {
      "epoch": 0.8808510638297873,
      "grad_norm": 0.80455292594799,
      "learning_rate": 8.90915741234015e-06,
      "loss": 0.8639,
      "step": 69
    },
    {
      "epoch": 0.8936170212765957,
      "grad_norm": 0.798116367786634,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.8419,
      "step": 70
    },
    {
      "epoch": 0.9063829787234042,
      "grad_norm": 0.6175272732760302,
      "learning_rate": 8.814147859311333e-06,
      "loss": 0.8281,
      "step": 71
    },
    {
      "epoch": 0.9191489361702128,
      "grad_norm": 0.6991366864152418,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.7673,
      "step": 72
    },
    {
      "epoch": 0.9319148936170213,
      "grad_norm": 0.7800616049786662,
      "learning_rate": 8.715724127386971e-06,
      "loss": 0.8352,
      "step": 73
    },
    {
      "epoch": 0.9446808510638298,
      "grad_norm": 0.6071910134631383,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.7964,
      "step": 74
    },
    {
      "epoch": 0.9574468085106383,
      "grad_norm": 0.7175207409525108,
      "learning_rate": 8.613974319136959e-06,
      "loss": 0.8003,
      "step": 75
    },
    {
      "epoch": 0.9702127659574468,
      "grad_norm": 0.7269990693905276,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.8035,
      "step": 76
    },
    {
      "epoch": 0.9829787234042553,
      "grad_norm": 0.643014711354184,
      "learning_rate": 8.508989514419959e-06,
      "loss": 0.832,
      "step": 77
    },
    {
      "epoch": 0.9957446808510638,
      "grad_norm": 0.5906253178794993,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.8212,
      "step": 78
    },
    {
      "epoch": 1.0085106382978724,
      "grad_norm": 1.4253189700474573,
      "learning_rate": 8.400863688854598e-06,
      "loss": 1.3589,
      "step": 79
    },
    {
      "epoch": 1.0212765957446808,
      "grad_norm": 0.6252290101034432,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.681,
      "step": 80
    },
    {
      "epoch": 1.0340425531914894,
      "grad_norm": 0.6126983062781691,
      "learning_rate": 8.289693629698564e-06,
      "loss": 0.7441,
      "step": 81
    },
    {
      "epoch": 1.0468085106382978,
      "grad_norm": 0.5764424319094782,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.743,
      "step": 82
    },
    {
      "epoch": 1.0595744680851065,
      "grad_norm": 0.6951973478221167,
      "learning_rate": 8.175578849210894e-06,
      "loss": 0.8549,
      "step": 83
    },
    {
      "epoch": 1.0723404255319149,
      "grad_norm": 0.741475138432806,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.7314,
      "step": 84
    },
    {
      "epoch": 1.0851063829787233,
      "grad_norm": 0.619112834273637,
      "learning_rate": 8.058621495575032e-06,
      "loss": 0.69,
      "step": 85
    },
    {
      "epoch": 1.097872340425532,
      "grad_norm": 0.6709089764707235,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.7606,
      "step": 86
    },
    {
      "epoch": 1.1106382978723404,
      "grad_norm": 0.9095717040896454,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.7677,
      "step": 87
    },
    {
      "epoch": 1.123404255319149,
      "grad_norm": 0.7296980623269048,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.7596,
      "step": 88
    },
    {
      "epoch": 1.1361702127659574,
      "grad_norm": 0.775655214973485,
      "learning_rate": 7.81660029031811e-06,
      "loss": 0.7416,
      "step": 89
    },
    {
      "epoch": 1.148936170212766,
      "grad_norm": 0.6743944889465332,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.688,
      "step": 90
    },
    {
      "epoch": 1.1617021276595745,
      "grad_norm": 0.6335100743206861,
      "learning_rate": 7.691753080453413e-06,
      "loss": 0.7427,
      "step": 91
    },
    {
      "epoch": 1.174468085106383,
      "grad_norm": 0.711181947370624,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.7761,
      "step": 92
    },
    {
      "epoch": 1.1872340425531915,
      "grad_norm": 0.7514220496808613,
      "learning_rate": 7.564496387029532e-06,
      "loss": 0.8177,
      "step": 93
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.658573568051986,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.731,
      "step": 94
    },
    {
      "epoch": 1.2127659574468086,
      "grad_norm": 0.6461429813659317,
      "learning_rate": 7.434944122021837e-06,
      "loss": 0.7076,
      "step": 95
    },
    {
      "epoch": 1.225531914893617,
      "grad_norm": 0.6072398670230126,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.7707,
      "step": 96
    },
    {
      "epoch": 1.2382978723404254,
      "grad_norm": 0.5683579113242971,
      "learning_rate": 7.303212252253163e-06,
      "loss": 0.7246,
      "step": 97
    },
    {
      "epoch": 1.251063829787234,
      "grad_norm": 0.6470968617096235,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.8116,
      "step": 98
    },
    {
      "epoch": 1.2638297872340425,
      "grad_norm": 0.6290836574482417,
      "learning_rate": 7.169418695587791e-06,
      "loss": 0.7499,
      "step": 99
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 0.5386462029087553,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.7554,
      "step": 100
    },
    {
      "epoch": 1.2893617021276595,
      "grad_norm": 0.6379302400766887,
      "learning_rate": 7.033683215379002e-06,
      "loss": 0.6596,
      "step": 101
    },
    {
      "epoch": 1.302127659574468,
      "grad_norm": 0.6738218338500571,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.8226,
      "step": 102
    },
    {
      "epoch": 1.3148936170212766,
      "grad_norm": 0.592967066545746,
      "learning_rate": 6.896127313264643e-06,
      "loss": 0.8034,
      "step": 103
    },
    {
      "epoch": 1.327659574468085,
      "grad_norm": 0.5714401976787592,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.7348,
      "step": 104
    },
    {
      "epoch": 1.3404255319148937,
      "grad_norm": 0.5976528961528519,
      "learning_rate": 6.7568741204067145e-06,
      "loss": 0.7203,
      "step": 105
    },
    {
      "epoch": 1.353191489361702,
      "grad_norm": 0.5777615361976374,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.7419,
      "step": 106
    },
    {
      "epoch": 1.3659574468085105,
      "grad_norm": 0.5348666700952892,
      "learning_rate": 6.616048287272301e-06,
      "loss": 0.7898,
      "step": 107
    },
    {
      "epoch": 1.3787234042553191,
      "grad_norm": 0.6621307601091048,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.8062,
      "step": 108
    },
    {
      "epoch": 1.3914893617021278,
      "grad_norm": 0.691721594153437,
      "learning_rate": 6.473775872054522e-06,
      "loss": 0.8034,
      "step": 109
    },
    {
      "epoch": 1.4042553191489362,
      "grad_norm": 0.5686130190581552,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.7673,
      "step": 110
    },
    {
      "epoch": 1.4170212765957446,
      "grad_norm": 0.4756378346544592,
      "learning_rate": 6.330184227833376e-06,
      "loss": 0.6315,
      "step": 111
    },
    {
      "epoch": 1.4297872340425533,
      "grad_norm": 0.6206136672748521,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.7698,
      "step": 112
    },
    {
      "epoch": 1.4425531914893617,
      "grad_norm": 0.6303389185178494,
      "learning_rate": 6.185401888577488e-06,
      "loss": 0.8267,
      "step": 113
    },
    {
      "epoch": 1.4553191489361703,
      "grad_norm": 0.5452107765410494,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.6971,
      "step": 114
    },
    {
      "epoch": 1.4680851063829787,
      "grad_norm": 0.6060781455714925,
      "learning_rate": 6.039558454088796e-06,
      "loss": 0.8687,
      "step": 115
    },
    {
      "epoch": 1.4808510638297872,
      "grad_norm": 0.6013851352507967,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.732,
      "step": 116
    },
    {
      "epoch": 1.4936170212765958,
      "grad_norm": 0.5562358332246483,
      "learning_rate": 5.892784473993184e-06,
      "loss": 0.6502,
      "step": 117
    },
    {
      "epoch": 1.5063829787234042,
      "grad_norm": 0.5675479514156825,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.7959,
      "step": 118
    },
    {
      "epoch": 1.5191489361702128,
      "grad_norm": 0.546953474852278,
      "learning_rate": 5.745211330880872e-06,
      "loss": 0.7351,
      "step": 119
    },
    {
      "epoch": 1.5319148936170213,
      "grad_norm": 0.5583886944268567,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.7528,
      "step": 120
    },
    {
      "epoch": 1.5446808510638297,
      "grad_norm": 0.5600086311533115,
      "learning_rate": 5.596971122701221e-06,
      "loss": 0.8955,
      "step": 121
    },
    {
      "epoch": 1.5574468085106383,
      "grad_norm": 0.6029023748669958,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.7261,
      "step": 122
    },
    {
      "epoch": 1.570212765957447,
      "grad_norm": 0.6342694584033263,
      "learning_rate": 5.448196544517168e-06,
      "loss": 0.7679,
      "step": 123
    },
    {
      "epoch": 1.5829787234042554,
      "grad_norm": 0.6065847834649665,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.7638,
      "step": 124
    },
    {
      "epoch": 1.5957446808510638,
      "grad_norm": 0.5265674527652279,
      "learning_rate": 5.299020769725172e-06,
      "loss": 0.787,
      "step": 125
    },
    {
      "epoch": 1.6085106382978722,
      "grad_norm": 0.6149521469743251,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.8122,
      "step": 126
    },
    {
      "epoch": 1.6212765957446809,
      "grad_norm": 0.6086968303307506,
      "learning_rate": 5.1495773308469935e-06,
      "loss": 0.7534,
      "step": 127
    },
    {
      "epoch": 1.6340425531914895,
      "grad_norm": 0.5682726195284596,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.7181,
      "step": 128
    },
    {
      "epoch": 1.646808510638298,
      "grad_norm": 0.5754891009333322,
      "learning_rate": 5e-06,
      "loss": 0.892,
      "step": 129
    },
    {
      "epoch": 1.6595744680851063,
      "grad_norm": 0.6043646952027831,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.7209,
      "step": 130
    },
    {
      "epoch": 1.6723404255319148,
      "grad_norm": 0.5578664828925858,
      "learning_rate": 4.850422669153009e-06,
      "loss": 0.7289,
      "step": 131
    },
    {
      "epoch": 1.6851063829787234,
      "grad_norm": 0.5988474734097357,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.7493,
      "step": 132
    },
    {
      "epoch": 1.697872340425532,
      "grad_norm": 0.5279090391793908,
      "learning_rate": 4.700979230274829e-06,
      "loss": 0.6385,
      "step": 133
    },
    {
      "epoch": 1.7106382978723405,
      "grad_norm": 0.6477329260959007,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.868,
      "step": 134
    },
    {
      "epoch": 1.7234042553191489,
      "grad_norm": 0.574998428834536,
      "learning_rate": 4.551803455482833e-06,
      "loss": 0.7627,
      "step": 135
    },
    {
      "epoch": 1.7361702127659573,
      "grad_norm": 0.5203426261066783,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.7253,
      "step": 136
    },
    {
      "epoch": 1.748936170212766,
      "grad_norm": 0.6267633363557024,
      "learning_rate": 4.4030288772987795e-06,
      "loss": 0.7586,
      "step": 137
    },
    {
      "epoch": 1.7617021276595746,
      "grad_norm": 0.5632632149868597,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.808,
      "step": 138
    },
    {
      "epoch": 1.774468085106383,
      "grad_norm": 0.5630999443363466,
      "learning_rate": 4.254788669119127e-06,
      "loss": 0.7702,
      "step": 139
    },
    {
      "epoch": 1.7872340425531914,
      "grad_norm": 0.5442195790445595,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.7557,
      "step": 140
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5883967505552894,
      "learning_rate": 4.107215526006818e-06,
      "loss": 0.7419,
      "step": 141
    },
    {
      "epoch": 1.8127659574468085,
      "grad_norm": 0.6080687924951912,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.666,
      "step": 142
    },
    {
      "epoch": 1.825531914893617,
      "grad_norm": 0.6251220871700776,
      "learning_rate": 3.960441545911205e-06,
      "loss": 0.772,
      "step": 143
    },
    {
      "epoch": 1.8382978723404255,
      "grad_norm": 0.5777374888887631,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.7972,
      "step": 144
    },
    {
      "epoch": 1.851063829787234,
      "grad_norm": 0.5341946785783167,
      "learning_rate": 3.8145981114225135e-06,
      "loss": 0.6953,
      "step": 145
    },
    {
      "epoch": 1.8638297872340426,
      "grad_norm": 0.6516092653149019,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.8309,
      "step": 146
    },
    {
      "epoch": 1.8765957446808512,
      "grad_norm": 0.6366696504215543,
      "learning_rate": 3.669815772166625e-06,
      "loss": 0.6375,
      "step": 147
    },
    {
      "epoch": 1.8893617021276596,
      "grad_norm": 0.5764338420537471,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.7718,
      "step": 148
    },
    {
      "epoch": 1.902127659574468,
      "grad_norm": 0.5034839691318637,
      "learning_rate": 3.526224127945479e-06,
      "loss": 0.7586,
      "step": 149
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 0.659371553780018,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.8194,
      "step": 150
    },
    {
      "epoch": 1.9276595744680851,
      "grad_norm": 0.575488935335163,
      "learning_rate": 3.383951712727701e-06,
      "loss": 0.7998,
      "step": 151
    },
    {
      "epoch": 1.9404255319148938,
      "grad_norm": 0.6021998191038525,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.7879,
      "step": 152
    },
    {
      "epoch": 1.9531914893617022,
      "grad_norm": 0.5477249938045327,
      "learning_rate": 3.2431258795932863e-06,
      "loss": 0.7206,
      "step": 153
    },
    {
      "epoch": 1.9659574468085106,
      "grad_norm": 0.5811445369216853,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.7528,
      "step": 154
    },
    {
      "epoch": 1.978723404255319,
      "grad_norm": 0.5280702510308407,
      "learning_rate": 3.1038726867353587e-06,
      "loss": 0.7817,
      "step": 155
    },
    {
      "epoch": 1.9914893617021276,
      "grad_norm": 0.5652222500582458,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.7491,
      "step": 156
    },
    {
      "epoch": 2.0042553191489363,
      "grad_norm": 1.0505257444754954,
      "learning_rate": 2.966316784621e-06,
      "loss": 1.2112,
      "step": 157
    },
    {
      "epoch": 2.0170212765957447,
      "grad_norm": 0.5534679233773416,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.6694,
      "step": 158
    },
    {
      "epoch": 2.029787234042553,
      "grad_norm": 0.6416084502104811,
      "learning_rate": 2.83058130441221e-06,
      "loss": 0.7556,
      "step": 159
    },
    {
      "epoch": 2.0425531914893615,
      "grad_norm": 0.4798010455035571,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.6236,
      "step": 160
    },
    {
      "epoch": 2.0553191489361704,
      "grad_norm": 0.6989318235437988,
      "learning_rate": 2.6967877477468394e-06,
      "loss": 0.8026,
      "step": 161
    },
    {
      "epoch": 2.068085106382979,
      "grad_norm": 0.6281981748050607,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.693,
      "step": 162
    },
    {
      "epoch": 2.0808510638297872,
      "grad_norm": 0.5152717911140114,
      "learning_rate": 2.5650558779781635e-06,
      "loss": 0.695,
      "step": 163
    },
    {
      "epoch": 2.0936170212765957,
      "grad_norm": 0.5157593106953199,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.6958,
      "step": 164
    },
    {
      "epoch": 2.106382978723404,
      "grad_norm": 0.542890735627848,
      "learning_rate": 2.43550361297047e-06,
      "loss": 0.6422,
      "step": 165
    },
    {
      "epoch": 2.119148936170213,
      "grad_norm": 0.6354916831539539,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.7221,
      "step": 166
    },
    {
      "epoch": 2.1319148936170214,
      "grad_norm": 0.5071694106301671,
      "learning_rate": 2.3082469195465893e-06,
      "loss": 0.6881,
      "step": 167
    },
    {
      "epoch": 2.1446808510638298,
      "grad_norm": 0.5270976651431593,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.714,
      "step": 168
    },
    {
      "epoch": 2.157446808510638,
      "grad_norm": 0.6107743680203847,
      "learning_rate": 2.1833997096818897e-06,
      "loss": 0.8456,
      "step": 169
    },
    {
      "epoch": 2.1702127659574466,
      "grad_norm": 0.5394700063686253,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.6249,
      "step": 170
    },
    {
      "epoch": 2.1829787234042555,
      "grad_norm": 0.5667817318976859,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.7802,
      "step": 171
    },
    {
      "epoch": 2.195744680851064,
      "grad_norm": 0.5044529000484788,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.5842,
      "step": 172
    },
    {
      "epoch": 2.2085106382978723,
      "grad_norm": 0.6094419863003621,
      "learning_rate": 1.941378504424968e-06,
      "loss": 0.8071,
      "step": 173
    },
    {
      "epoch": 2.2212765957446807,
      "grad_norm": 0.505840415405961,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.7655,
      "step": 174
    },
    {
      "epoch": 2.2340425531914896,
      "grad_norm": 0.5329780431110791,
      "learning_rate": 1.8244211507891064e-06,
      "loss": 0.6509,
      "step": 175
    },
    {
      "epoch": 2.246808510638298,
      "grad_norm": 0.5560722646167355,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.6761,
      "step": 176
    },
    {
      "epoch": 2.2595744680851064,
      "grad_norm": 0.5278556863776027,
      "learning_rate": 1.7103063703014372e-06,
      "loss": 0.752,
      "step": 177
    },
    {
      "epoch": 2.272340425531915,
      "grad_norm": 0.5540546065487634,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.7025,
      "step": 178
    },
    {
      "epoch": 2.2851063829787233,
      "grad_norm": 0.5313374180186387,
      "learning_rate": 1.5991363111454023e-06,
      "loss": 0.6167,
      "step": 179
    },
    {
      "epoch": 2.297872340425532,
      "grad_norm": 0.562448332468006,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.5999,
      "step": 180
    },
    {
      "epoch": 2.3106382978723405,
      "grad_norm": 0.5214892741988327,
      "learning_rate": 1.4910104855800429e-06,
      "loss": 0.7743,
      "step": 181
    },
    {
      "epoch": 2.323404255319149,
      "grad_norm": 0.5174189126938886,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.735,
      "step": 182
    },
    {
      "epoch": 2.3361702127659574,
      "grad_norm": 0.45428482617916605,
      "learning_rate": 1.3860256808630429e-06,
      "loss": 0.6105,
      "step": 183
    },
    {
      "epoch": 2.348936170212766,
      "grad_norm": 0.5152662327467596,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.7582,
      "step": 184
    },
    {
      "epoch": 2.3617021276595747,
      "grad_norm": 0.5469842875723849,
      "learning_rate": 1.2842758726130283e-06,
      "loss": 0.8307,
      "step": 185
    },
    {
      "epoch": 2.374468085106383,
      "grad_norm": 0.465872379406897,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.6998,
      "step": 186
    },
    {
      "epoch": 2.3872340425531915,
      "grad_norm": 0.5505647045899656,
      "learning_rate": 1.1858521406886674e-06,
      "loss": 0.8048,
      "step": 187
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.5063316847953139,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.6056,
      "step": 188
    },
    {
      "epoch": 2.4127659574468083,
      "grad_norm": 0.50130245907352,
      "learning_rate": 1.0908425876598512e-06,
      "loss": 0.6458,
      "step": 189
    },
    {
      "epoch": 2.425531914893617,
      "grad_norm": 0.5208231184042346,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.7236,
      "step": 190
    },
    {
      "epoch": 2.4382978723404256,
      "grad_norm": 0.5030678286204384,
      "learning_rate": 9.993322599439692e-07,
      "loss": 0.688,
      "step": 191
    },
    {
      "epoch": 2.451063829787234,
      "grad_norm": 0.5019824022121786,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.7615,
      "step": 192
    },
    {
      "epoch": 2.4638297872340424,
      "grad_norm": 0.5156242724008653,
      "learning_rate": 9.114030716778433e-07,
      "loss": 0.6917,
      "step": 193
    },
    {
      "epoch": 2.476595744680851,
      "grad_norm": 0.4810684332662468,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.7123,
      "step": 194
    },
    {
      "epoch": 2.4893617021276597,
      "grad_norm": 0.49610188906380476,
      "learning_rate": 8.271337313934869e-07,
      "loss": 0.6604,
      "step": 195
    },
    {
      "epoch": 2.502127659574468,
      "grad_norm": 0.5096131733768843,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.7049,
      "step": 196
    },
    {
      "epoch": 2.5148936170212766,
      "grad_norm": 0.46084306907290395,
      "learning_rate": 7.465996715633028e-07,
      "loss": 0.6966,
      "step": 197
    },
    {
      "epoch": 2.527659574468085,
      "grad_norm": 0.49577140692888255,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.7798,
      "step": 198
    },
    {
      "epoch": 2.5404255319148934,
      "grad_norm": 0.5012719167353504,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.7255,
      "step": 199
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 0.43751384195216086,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.6144,
      "step": 200
    },
    {
      "epoch": 2.5659574468085107,
      "grad_norm": 0.4906859931725163,
      "learning_rate": 5.9702234071631e-07,
      "loss": 0.7007,
      "step": 201
    },
    {
      "epoch": 2.578723404255319,
      "grad_norm": 0.5031899446810962,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.6856,
      "step": 202
    },
    {
      "epoch": 2.5914893617021275,
      "grad_norm": 0.47867968621934653,
      "learning_rate": 5.281129616683167e-07,
      "loss": 0.6839,
      "step": 203
    },
    {
      "epoch": 2.604255319148936,
      "grad_norm": 0.5157873816638013,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.7011,
      "step": 204
    },
    {
      "epoch": 2.617021276595745,
      "grad_norm": 0.47874359226187657,
      "learning_rate": 4.632065271606756e-07,
      "loss": 0.7438,
      "step": 205
    },
    {
      "epoch": 2.629787234042553,
      "grad_norm": 0.46097043391449666,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.6271,
      "step": 206
    },
    {
      "epoch": 2.6425531914893616,
      "grad_norm": 0.497047485208018,
      "learning_rate": 4.0236113724274716e-07,
      "loss": 0.7694,
      "step": 207
    },
    {
      "epoch": 2.65531914893617,
      "grad_norm": 0.47599364906666225,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.8376,
      "step": 208
    },
    {
      "epoch": 2.6680851063829785,
      "grad_norm": 0.43864910005965657,
      "learning_rate": 3.4563125677897936e-07,
      "loss": 0.6243,
      "step": 209
    },
    {
      "epoch": 2.6808510638297873,
      "grad_norm": 0.4621825463897554,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.7003,
      "step": 210
    },
    {
      "epoch": 2.6936170212765957,
      "grad_norm": 0.48232343197094046,
      "learning_rate": 2.930676666954846e-07,
      "loss": 0.7206,
      "step": 211
    },
    {
      "epoch": 2.706382978723404,
      "grad_norm": 0.47104350651465504,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.6196,
      "step": 212
    },
    {
      "epoch": 2.719148936170213,
      "grad_norm": 0.5273694728646362,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.7903,
      "step": 213
    },
    {
      "epoch": 2.731914893617021,
      "grad_norm": 0.4826068178110643,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.6996,
      "step": 214
    },
    {
      "epoch": 2.74468085106383,
      "grad_norm": 0.49633999113379995,
      "learning_rate": 2.006237922855553e-07,
      "loss": 0.706,
      "step": 215
    },
    {
      "epoch": 2.7574468085106383,
      "grad_norm": 0.5169749746813925,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.6336,
      "step": 216
    },
    {
      "epoch": 2.7702127659574467,
      "grad_norm": 0.47272884332119,
      "learning_rate": 1.6082625774666793e-07,
      "loss": 0.6656,
      "step": 217
    },
    {
      "epoch": 2.7829787234042556,
      "grad_norm": 0.46057432003620846,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.7266,
      "step": 218
    },
    {
      "epoch": 2.795744680851064,
      "grad_norm": 0.4893287165876055,
      "learning_rate": 1.253604390908819e-07,
      "loss": 0.6562,
      "step": 219
    },
    {
      "epoch": 2.8085106382978724,
      "grad_norm": 0.4648250703865466,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.6747,
      "step": 220
    },
    {
      "epoch": 2.821276595744681,
      "grad_norm": 0.4968226962990124,
      "learning_rate": 9.42580830291373e-08,
      "loss": 0.6722,
      "step": 221
    },
    {
      "epoch": 2.8340425531914892,
      "grad_norm": 0.515324253421338,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.7334,
      "step": 222
    },
    {
      "epoch": 2.846808510638298,
      "grad_norm": 0.46320007701441,
      "learning_rate": 6.75470303823933e-08,
      "loss": 0.6765,
      "step": 223
    },
    {
      "epoch": 2.8595744680851065,
      "grad_norm": 0.48508235718645204,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.6531,
      "step": 224
    },
    {
      "epoch": 2.872340425531915,
      "grad_norm": 0.5147688482752547,
      "learning_rate": 4.52511911603265e-08,
      "loss": 0.7047,
      "step": 225
    },
    {
      "epoch": 2.8851063829787233,
      "grad_norm": 0.4629650700246944,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.7151,
      "step": 226
    },
    {
      "epoch": 2.8978723404255318,
      "grad_norm": 0.5345400043641719,
      "learning_rate": 2.7390523158633552e-08,
      "loss": 0.7742,
      "step": 227
    },
    {
      "epoch": 2.9106382978723406,
      "grad_norm": 0.42658862838231876,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.6379,
      "step": 228
    },
    {
      "epoch": 2.923404255319149,
      "grad_norm": 0.4761085131478779,
      "learning_rate": 1.3981014094099354e-08,
      "loss": 0.7514,
      "step": 229
    },
    {
      "epoch": 2.9361702127659575,
      "grad_norm": 0.4659539884115577,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.7313,
      "step": 230
    },
    {
      "epoch": 2.948936170212766,
      "grad_norm": 0.42768655407061484,
      "learning_rate": 5.034667293427053e-09,
      "loss": 0.5988,
      "step": 231
    },
    {
      "epoch": 2.9617021276595743,
      "grad_norm": 0.5010147150356298,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.7718,
      "step": 232
    },
    {
      "epoch": 2.974468085106383,
      "grad_norm": 0.47358962083493916,
      "learning_rate": 5.594909486328348e-10,
      "loss": 0.6915,
      "step": 233
    },
    {
      "epoch": 2.9872340425531916,
      "grad_norm": 0.4838276929973422,
      "learning_rate": 0.0,
      "loss": 0.7312,
      "step": 234
    },
    {
      "epoch": 2.9872340425531916,
      "step": 234,
      "total_flos": 180145845731328.0,
      "train_loss": 0.7942904473369957,
      "train_runtime": 3536.2592,
      "train_samples_per_second": 6.363,
      "train_steps_per_second": 0.066
    }
  ],
  "logging_steps": 1,
  "max_steps": 234,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 180145845731328.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}