{
  "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": 6.433622804272975,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 1.2027,
      "step": 1
    },
    {
      "epoch": 0.02553191489361702,
      "grad_norm": 6.406146610066624,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.1984,
      "step": 2
    },
    {
      "epoch": 0.03829787234042553,
      "grad_norm": 6.369133763179086,
      "learning_rate": 1.25e-06,
      "loss": 1.1839,
      "step": 3
    },
    {
      "epoch": 0.05106382978723404,
      "grad_norm": 6.239688951706421,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.2008,
      "step": 4
    },
    {
      "epoch": 0.06382978723404255,
      "grad_norm": 5.725473695055957,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 1.1594,
      "step": 5
    },
    {
      "epoch": 0.07659574468085106,
      "grad_norm": 4.830315030136056,
      "learning_rate": 2.5e-06,
      "loss": 1.1631,
      "step": 6
    },
    {
      "epoch": 0.08936170212765958,
      "grad_norm": 4.369137536951095,
      "learning_rate": 2.916666666666667e-06,
      "loss": 1.1432,
      "step": 7
    },
    {
      "epoch": 0.10212765957446808,
      "grad_norm": 2.72542957762778,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0909,
      "step": 8
    },
    {
      "epoch": 0.1148936170212766,
      "grad_norm": 2.4979340740744864,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 1.0602,
      "step": 9
    },
    {
      "epoch": 0.1276595744680851,
      "grad_norm": 2.3755723671193927,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.065,
      "step": 10
    },
    {
      "epoch": 0.14042553191489363,
      "grad_norm": 4.2839335365862965,
      "learning_rate": 4.583333333333333e-06,
      "loss": 1.0832,
      "step": 11
    },
    {
      "epoch": 0.15319148936170213,
      "grad_norm": 4.116111912140754,
      "learning_rate": 5e-06,
      "loss": 1.0428,
      "step": 12
    },
    {
      "epoch": 0.16595744680851063,
      "grad_norm": 4.005685697811548,
      "learning_rate": 5.416666666666667e-06,
      "loss": 1.0284,
      "step": 13
    },
    {
      "epoch": 0.17872340425531916,
      "grad_norm": 3.223331293112425,
      "learning_rate": 5.833333333333334e-06,
      "loss": 1.0322,
      "step": 14
    },
    {
      "epoch": 0.19148936170212766,
      "grad_norm": 3.373013206868157,
      "learning_rate": 6.25e-06,
      "loss": 1.0332,
      "step": 15
    },
    {
      "epoch": 0.20425531914893616,
      "grad_norm": 2.720673997841046,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.941,
      "step": 16
    },
    {
      "epoch": 0.2170212765957447,
      "grad_norm": 2.3602360691078172,
      "learning_rate": 7.083333333333335e-06,
      "loss": 0.9753,
      "step": 17
    },
    {
      "epoch": 0.2297872340425532,
      "grad_norm": 2.123106209821333,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9487,
      "step": 18
    },
    {
      "epoch": 0.2425531914893617,
      "grad_norm": 1.9375668256870149,
      "learning_rate": 7.916666666666667e-06,
      "loss": 0.9297,
      "step": 19
    },
    {
      "epoch": 0.2553191489361702,
      "grad_norm": 1.679662940193537,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9346,
      "step": 20
    },
    {
      "epoch": 0.2680851063829787,
      "grad_norm": 1.5536294679138145,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.9435,
      "step": 21
    },
    {
      "epoch": 0.28085106382978725,
      "grad_norm": 1.5959235415078292,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.9159,
      "step": 22
    },
    {
      "epoch": 0.2936170212765957,
      "grad_norm": 1.3251855095591336,
      "learning_rate": 9.583333333333335e-06,
      "loss": 0.924,
      "step": 23
    },
    {
      "epoch": 0.30638297872340425,
      "grad_norm": 1.333931250234237,
      "learning_rate": 1e-05,
      "loss": 0.9087,
      "step": 24
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 1.0631172254893704,
      "learning_rate": 9.999440509051367e-06,
      "loss": 0.8619,
      "step": 25
    },
    {
      "epoch": 0.33191489361702126,
      "grad_norm": 1.1428090130087067,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.8663,
      "step": 26
    },
    {
      "epoch": 0.3446808510638298,
      "grad_norm": 0.9408863721323562,
      "learning_rate": 9.994965332706574e-06,
      "loss": 0.9178,
      "step": 27
    },
    {
      "epoch": 0.3574468085106383,
      "grad_norm": 0.8661346108852108,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.8742,
      "step": 28
    },
    {
      "epoch": 0.3702127659574468,
      "grad_norm": 1.0870551192596247,
      "learning_rate": 9.986018985905901e-06,
      "loss": 0.9193,
      "step": 29
    },
    {
      "epoch": 0.3829787234042553,
      "grad_norm": 0.8726063330094801,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.8831,
      "step": 30
    },
    {
      "epoch": 0.39574468085106385,
      "grad_norm": 0.7897053030050917,
      "learning_rate": 9.972609476841368e-06,
      "loss": 0.8744,
      "step": 31
    },
    {
      "epoch": 0.4085106382978723,
      "grad_norm": 0.8532203062862161,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.8444,
      "step": 32
    },
    {
      "epoch": 0.42127659574468085,
      "grad_norm": 0.9015264429234872,
      "learning_rate": 9.954748808839675e-06,
      "loss": 0.9029,
      "step": 33
    },
    {
      "epoch": 0.4340425531914894,
      "grad_norm": 0.9314825609693578,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.906,
      "step": 34
    },
    {
      "epoch": 0.44680851063829785,
      "grad_norm": 0.8857778600988849,
      "learning_rate": 9.932452969617607e-06,
      "loss": 0.8727,
      "step": 35
    },
    {
      "epoch": 0.4595744680851064,
      "grad_norm": 0.9102922225453883,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.8703,
      "step": 36
    },
    {
      "epoch": 0.4723404255319149,
      "grad_norm": 0.9373330088387606,
      "learning_rate": 9.905741916970863e-06,
      "loss": 0.8632,
      "step": 37
    },
    {
      "epoch": 0.4851063829787234,
      "grad_norm": 0.648113277413392,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.8709,
      "step": 38
    },
    {
      "epoch": 0.4978723404255319,
      "grad_norm": 0.8476328265586036,
      "learning_rate": 9.874639560909118e-06,
      "loss": 0.8641,
      "step": 39
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 0.9530645465852795,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.8873,
      "step": 40
    },
    {
      "epoch": 0.5234042553191489,
      "grad_norm": 0.8834502637259801,
      "learning_rate": 9.839173742253334e-06,
      "loss": 0.9088,
      "step": 41
    },
    {
      "epoch": 0.5361702127659574,
      "grad_norm": 0.6674564507015708,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.8836,
      "step": 42
    },
    {
      "epoch": 0.548936170212766,
      "grad_norm": 0.89233659474358,
      "learning_rate": 9.799376207714446e-06,
      "loss": 0.861,
      "step": 43
    },
    {
      "epoch": 0.5617021276595745,
      "grad_norm": 0.9657269278077453,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.8508,
      "step": 44
    },
    {
      "epoch": 0.574468085106383,
      "grad_norm": 0.8180964670581452,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.8545,
      "step": 45
    },
    {
      "epoch": 0.5872340425531914,
      "grad_norm": 0.6255777011672562,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.8117,
      "step": 46
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.7667163975506532,
      "learning_rate": 9.706932333304518e-06,
      "loss": 0.8895,
      "step": 47
    },
    {
      "epoch": 0.6127659574468085,
      "grad_norm": 0.8020605434247621,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.854,
      "step": 48
    },
    {
      "epoch": 0.625531914893617,
      "grad_norm": 0.7153116420784209,
      "learning_rate": 9.654368743221022e-06,
      "loss": 0.8491,
      "step": 49
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 0.7632863572263763,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.8296,
      "step": 50
    },
    {
      "epoch": 0.6510638297872341,
      "grad_norm": 0.7156700482115346,
      "learning_rate": 9.597638862757255e-06,
      "loss": 0.8269,
      "step": 51
    },
    {
      "epoch": 0.6638297872340425,
      "grad_norm": 0.7419920177521214,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.8732,
      "step": 52
    },
    {
      "epoch": 0.676595744680851,
      "grad_norm": 0.7274084462685031,
      "learning_rate": 9.536793472839325e-06,
      "loss": 0.8571,
      "step": 53
    },
    {
      "epoch": 0.6893617021276596,
      "grad_norm": 0.8071133486525568,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.8618,
      "step": 54
    },
    {
      "epoch": 0.7021276595744681,
      "grad_norm": 0.7040439538018551,
      "learning_rate": 9.471887038331686e-06,
      "loss": 0.8062,
      "step": 55
    },
    {
      "epoch": 0.7148936170212766,
      "grad_norm": 0.6461644492867274,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.8151,
      "step": 56
    },
    {
      "epoch": 0.7276595744680852,
      "grad_norm": 0.7749362259461846,
      "learning_rate": 9.40297765928369e-06,
      "loss": 0.8244,
      "step": 57
    },
    {
      "epoch": 0.7404255319148936,
      "grad_norm": 0.6824545661899104,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.8576,
      "step": 58
    },
    {
      "epoch": 0.7531914893617021,
      "grad_norm": 0.7432671437330083,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.8034,
      "step": 59
    },
    {
      "epoch": 0.7659574468085106,
      "grad_norm": 0.7600556786309642,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.8458,
      "step": 60
    },
    {
      "epoch": 0.7787234042553192,
      "grad_norm": 0.6265580676056108,
      "learning_rate": 9.253400328436699e-06,
      "loss": 0.861,
      "step": 61
    },
    {
      "epoch": 0.7914893617021277,
      "grad_norm": 0.6225560200815765,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.8604,
      "step": 62
    },
    {
      "epoch": 0.8042553191489362,
      "grad_norm": 0.8810350842991577,
      "learning_rate": 9.172866268606514e-06,
      "loss": 0.8427,
      "step": 63
    },
    {
      "epoch": 0.8170212765957446,
      "grad_norm": 0.6513841847105947,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.8028,
      "step": 64
    },
    {
      "epoch": 0.8297872340425532,
      "grad_norm": 0.6197128493703077,
      "learning_rate": 9.088596928322158e-06,
      "loss": 0.8656,
      "step": 65
    },
    {
      "epoch": 0.8425531914893617,
      "grad_norm": 0.6815619030192367,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.865,
      "step": 66
    },
    {
      "epoch": 0.8553191489361702,
      "grad_norm": 0.6559308467623148,
      "learning_rate": 9.000667740056033e-06,
      "loss": 0.8462,
      "step": 67
    },
    {
      "epoch": 0.8680851063829788,
      "grad_norm": 0.6623712768744643,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.819,
      "step": 68
    },
    {
      "epoch": 0.8808510638297873,
      "grad_norm": 0.6180928952593647,
      "learning_rate": 8.90915741234015e-06,
      "loss": 0.7912,
      "step": 69
    },
    {
      "epoch": 0.8936170212765957,
      "grad_norm": 0.6273034711494855,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.8346,
      "step": 70
    },
    {
      "epoch": 0.9063829787234042,
      "grad_norm": 0.5954230015766475,
      "learning_rate": 8.814147859311333e-06,
      "loss": 0.8285,
      "step": 71
    },
    {
      "epoch": 0.9191489361702128,
      "grad_norm": 0.6074571433084066,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.846,
      "step": 72
    },
    {
      "epoch": 0.9319148936170213,
      "grad_norm": 0.6369989484635978,
      "learning_rate": 8.715724127386971e-06,
      "loss": 0.8523,
      "step": 73
    },
    {
      "epoch": 0.9446808510638298,
      "grad_norm": 0.7075912076720394,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.8437,
      "step": 74
    },
    {
      "epoch": 0.9574468085106383,
      "grad_norm": 0.6088301978445659,
      "learning_rate": 8.613974319136959e-06,
      "loss": 0.8301,
      "step": 75
    },
    {
      "epoch": 0.9702127659574468,
      "grad_norm": 0.6095867804621813,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.8453,
      "step": 76
    },
    {
      "epoch": 0.9829787234042553,
      "grad_norm": 0.5939945482764697,
      "learning_rate": 8.508989514419959e-06,
      "loss": 0.8019,
      "step": 77
    },
    {
      "epoch": 0.9957446808510638,
      "grad_norm": 0.6748168678734431,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.842,
      "step": 78
    },
    {
      "epoch": 1.0085106382978724,
      "grad_norm": 1.547759173938074,
      "learning_rate": 8.400863688854598e-06,
      "loss": 1.3859,
      "step": 79
    },
    {
      "epoch": 1.0212765957446808,
      "grad_norm": 0.5834496082903455,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.6686,
      "step": 80
    },
    {
      "epoch": 1.0340425531914894,
      "grad_norm": 0.8717786274161273,
      "learning_rate": 8.289693629698564e-06,
      "loss": 0.8405,
      "step": 81
    },
    {
      "epoch": 1.0468085106382978,
      "grad_norm": 0.6157547818545464,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.82,
      "step": 82
    },
    {
      "epoch": 1.0595744680851065,
      "grad_norm": 0.7130510637018735,
      "learning_rate": 8.175578849210894e-06,
      "loss": 0.7519,
      "step": 83
    },
    {
      "epoch": 1.0723404255319149,
      "grad_norm": 0.690303000029393,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.751,
      "step": 84
    },
    {
      "epoch": 1.0851063829787233,
      "grad_norm": 0.7079577745236357,
      "learning_rate": 8.058621495575032e-06,
      "loss": 0.7809,
      "step": 85
    },
    {
      "epoch": 1.097872340425532,
      "grad_norm": 0.6652157947641348,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.7239,
      "step": 86
    },
    {
      "epoch": 1.1106382978723404,
      "grad_norm": 0.6609980037072015,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.7185,
      "step": 87
    },
    {
      "epoch": 1.123404255319149,
      "grad_norm": 0.6572439615475121,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.7618,
      "step": 88
    },
    {
      "epoch": 1.1361702127659574,
      "grad_norm": 0.6524069711521925,
      "learning_rate": 7.81660029031811e-06,
      "loss": 0.7565,
      "step": 89
    },
    {
      "epoch": 1.148936170212766,
      "grad_norm": 0.7267875020327975,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.7626,
      "step": 90
    },
    {
      "epoch": 1.1617021276595745,
      "grad_norm": 0.6412064880318965,
      "learning_rate": 7.691753080453413e-06,
      "loss": 0.7504,
      "step": 91
    },
    {
      "epoch": 1.174468085106383,
      "grad_norm": 0.6837901280994855,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.7695,
      "step": 92
    },
    {
      "epoch": 1.1872340425531915,
      "grad_norm": 0.6871789853177429,
      "learning_rate": 7.564496387029532e-06,
      "loss": 0.8094,
      "step": 93
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.6238936487998985,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7346,
      "step": 94
    },
    {
      "epoch": 1.2127659574468086,
      "grad_norm": 0.6150321617049529,
      "learning_rate": 7.434944122021837e-06,
      "loss": 0.8103,
      "step": 95
    },
    {
      "epoch": 1.225531914893617,
      "grad_norm": 0.7953351451984426,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.7658,
      "step": 96
    },
    {
      "epoch": 1.2382978723404254,
      "grad_norm": 0.6687467735647058,
      "learning_rate": 7.303212252253163e-06,
      "loss": 0.7674,
      "step": 97
    },
    {
      "epoch": 1.251063829787234,
      "grad_norm": 0.6041472836507435,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.7526,
      "step": 98
    },
    {
      "epoch": 1.2638297872340425,
      "grad_norm": 0.637437469544626,
      "learning_rate": 7.169418695587791e-06,
      "loss": 0.7459,
      "step": 99
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 0.6756564518794657,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.7433,
      "step": 100
    },
    {
      "epoch": 1.2893617021276595,
      "grad_norm": 0.617925149676544,
      "learning_rate": 7.033683215379002e-06,
      "loss": 0.8091,
      "step": 101
    },
    {
      "epoch": 1.302127659574468,
      "grad_norm": 0.6331398323341669,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.7685,
      "step": 102
    },
    {
      "epoch": 1.3148936170212766,
      "grad_norm": 0.6588781402724724,
      "learning_rate": 6.896127313264643e-06,
      "loss": 0.7512,
      "step": 103
    },
    {
      "epoch": 1.327659574468085,
      "grad_norm": 0.5567840865056657,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.7207,
      "step": 104
    },
    {
      "epoch": 1.3404255319148937,
      "grad_norm": 0.6450623261539175,
      "learning_rate": 6.7568741204067145e-06,
      "loss": 0.7743,
      "step": 105
    },
    {
      "epoch": 1.353191489361702,
      "grad_norm": 0.5959132949800964,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.7671,
      "step": 106
    },
    {
      "epoch": 1.3659574468085105,
      "grad_norm": 0.5845969623624405,
      "learning_rate": 6.616048287272301e-06,
      "loss": 0.6478,
      "step": 107
    },
    {
      "epoch": 1.3787234042553191,
      "grad_norm": 0.759276559832424,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.8928,
      "step": 108
    },
    {
      "epoch": 1.3914893617021278,
      "grad_norm": 0.656020228293016,
      "learning_rate": 6.473775872054522e-06,
      "loss": 0.8068,
      "step": 109
    },
    {
      "epoch": 1.4042553191489362,
      "grad_norm": 0.5760724781489475,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.6719,
      "step": 110
    },
    {
      "epoch": 1.4170212765957446,
      "grad_norm": 0.6909070411507484,
      "learning_rate": 6.330184227833376e-06,
      "loss": 0.7651,
      "step": 111
    },
    {
      "epoch": 1.4297872340425533,
      "grad_norm": 0.7063008528557656,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.8234,
      "step": 112
    },
    {
      "epoch": 1.4425531914893617,
      "grad_norm": 0.5299402625765229,
      "learning_rate": 6.185401888577488e-06,
      "loss": 0.7144,
      "step": 113
    },
    {
      "epoch": 1.4553191489361703,
      "grad_norm": 0.7426768197949125,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.8006,
      "step": 114
    },
    {
      "epoch": 1.4680851063829787,
      "grad_norm": 0.6609099273855246,
      "learning_rate": 6.039558454088796e-06,
      "loss": 0.69,
      "step": 115
    },
    {
      "epoch": 1.4808510638297872,
      "grad_norm": 0.7192549036207418,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.8544,
      "step": 116
    },
    {
      "epoch": 1.4936170212765958,
      "grad_norm": 0.5884339069720301,
      "learning_rate": 5.892784473993184e-06,
      "loss": 0.7449,
      "step": 117
    },
    {
      "epoch": 1.5063829787234042,
      "grad_norm": 0.7350852245605305,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.7141,
      "step": 118
    },
    {
      "epoch": 1.5191489361702128,
      "grad_norm": 0.6979695732369594,
      "learning_rate": 5.745211330880872e-06,
      "loss": 0.8249,
      "step": 119
    },
    {
      "epoch": 1.5319148936170213,
      "grad_norm": 0.5235553438939273,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.7202,
      "step": 120
    },
    {
      "epoch": 1.5446808510638297,
      "grad_norm": 0.682351171653146,
      "learning_rate": 5.596971122701221e-06,
      "loss": 0.7705,
      "step": 121
    },
    {
      "epoch": 1.5574468085106383,
      "grad_norm": 0.5620126307246519,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.7026,
      "step": 122
    },
    {
      "epoch": 1.570212765957447,
      "grad_norm": 0.5839082634964239,
      "learning_rate": 5.448196544517168e-06,
      "loss": 0.7979,
      "step": 123
    },
    {
      "epoch": 1.5829787234042554,
      "grad_norm": 0.5350002506195382,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.7143,
      "step": 124
    },
    {
      "epoch": 1.5957446808510638,
      "grad_norm": 0.6946599115440883,
      "learning_rate": 5.299020769725172e-06,
      "loss": 0.7873,
      "step": 125
    },
    {
      "epoch": 1.6085106382978722,
      "grad_norm": 0.5748638691181778,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.7365,
      "step": 126
    },
    {
      "epoch": 1.6212765957446809,
      "grad_norm": 0.6152731198473932,
      "learning_rate": 5.1495773308469935e-06,
      "loss": 0.8644,
      "step": 127
    },
    {
      "epoch": 1.6340425531914895,
      "grad_norm": 0.5381156315117113,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.645,
      "step": 128
    },
    {
      "epoch": 1.646808510638298,
      "grad_norm": 0.6120012511222306,
      "learning_rate": 5e-06,
      "loss": 0.7568,
      "step": 129
    },
    {
      "epoch": 1.6595744680851063,
      "grad_norm": 0.6206112452727576,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.7629,
      "step": 130
    },
    {
      "epoch": 1.6723404255319148,
      "grad_norm": 0.5874975559118444,
      "learning_rate": 4.850422669153009e-06,
      "loss": 0.7951,
      "step": 131
    },
    {
      "epoch": 1.6851063829787234,
      "grad_norm": 0.585196228318139,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.742,
      "step": 132
    },
    {
      "epoch": 1.697872340425532,
      "grad_norm": 0.6903440231119622,
      "learning_rate": 4.700979230274829e-06,
      "loss": 0.885,
      "step": 133
    },
    {
      "epoch": 1.7106382978723405,
      "grad_norm": 0.5667409120364741,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.6209,
      "step": 134
    },
    {
      "epoch": 1.7234042553191489,
      "grad_norm": 0.5497052698552617,
      "learning_rate": 4.551803455482833e-06,
      "loss": 0.7317,
      "step": 135
    },
    {
      "epoch": 1.7361702127659573,
      "grad_norm": 0.5684834730769993,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.7886,
      "step": 136
    },
    {
      "epoch": 1.748936170212766,
      "grad_norm": 0.6194533934703746,
      "learning_rate": 4.4030288772987795e-06,
      "loss": 0.6998,
      "step": 137
    },
    {
      "epoch": 1.7617021276595746,
      "grad_norm": 0.6080025275110034,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.7685,
      "step": 138
    },
    {
      "epoch": 1.774468085106383,
      "grad_norm": 0.6667437872795133,
      "learning_rate": 4.254788669119127e-06,
      "loss": 0.8504,
      "step": 139
    },
    {
      "epoch": 1.7872340425531914,
      "grad_norm": 0.4781266342412309,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.6402,
      "step": 140
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.6254395295795602,
      "learning_rate": 4.107215526006818e-06,
      "loss": 0.8257,
      "step": 141
    },
    {
      "epoch": 1.8127659574468085,
      "grad_norm": 0.6183136446673195,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.7743,
      "step": 142
    },
    {
      "epoch": 1.825531914893617,
      "grad_norm": 0.5212997216190336,
      "learning_rate": 3.960441545911205e-06,
      "loss": 0.6946,
      "step": 143
    },
    {
      "epoch": 1.8382978723404255,
      "grad_norm": 0.5098103809041011,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.7032,
      "step": 144
    },
    {
      "epoch": 1.851063829787234,
      "grad_norm": 0.5515247196630589,
      "learning_rate": 3.8145981114225135e-06,
      "loss": 0.7748,
      "step": 145
    },
    {
      "epoch": 1.8638297872340426,
      "grad_norm": 0.5394319928110296,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.8035,
      "step": 146
    },
    {
      "epoch": 1.8765957446808512,
      "grad_norm": 0.5389311901179236,
      "learning_rate": 3.669815772166625e-06,
      "loss": 0.7535,
      "step": 147
    },
    {
      "epoch": 1.8893617021276596,
      "grad_norm": 0.5019557383141358,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.7643,
      "step": 148
    },
    {
      "epoch": 1.902127659574468,
      "grad_norm": 0.5222037980369133,
      "learning_rate": 3.526224127945479e-06,
      "loss": 0.7609,
      "step": 149
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 0.504024959936413,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.7115,
      "step": 150
    },
    {
      "epoch": 1.9276595744680851,
      "grad_norm": 0.5265230348715444,
      "learning_rate": 3.383951712727701e-06,
      "loss": 0.858,
      "step": 151
    },
    {
      "epoch": 1.9404255319148938,
      "grad_norm": 0.5287078624508609,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.7184,
      "step": 152
    },
    {
      "epoch": 1.9531914893617022,
      "grad_norm": 0.5415420248763081,
      "learning_rate": 3.2431258795932863e-06,
      "loss": 0.7139,
      "step": 153
    },
    {
      "epoch": 1.9659574468085106,
      "grad_norm": 0.634512902103396,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.7809,
      "step": 154
    },
    {
      "epoch": 1.978723404255319,
      "grad_norm": 0.5487803141993715,
      "learning_rate": 3.1038726867353587e-06,
      "loss": 0.7562,
      "step": 155
    },
    {
      "epoch": 1.9914893617021276,
      "grad_norm": 0.6076957525111952,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.76,
      "step": 156
    },
    {
      "epoch": 2.0042553191489363,
      "grad_norm": 1.2666145516000928,
      "learning_rate": 2.966316784621e-06,
      "loss": 1.1093,
      "step": 157
    },
    {
      "epoch": 2.0170212765957447,
      "grad_norm": 0.5783416608389275,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.6804,
      "step": 158
    },
    {
      "epoch": 2.029787234042553,
      "grad_norm": 0.5339936797853089,
      "learning_rate": 2.83058130441221e-06,
      "loss": 0.7346,
      "step": 159
    },
    {
      "epoch": 2.0425531914893615,
      "grad_norm": 0.5121545171163273,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.7322,
      "step": 160
    },
    {
      "epoch": 2.0553191489361704,
      "grad_norm": 0.5688805304516626,
      "learning_rate": 2.6967877477468394e-06,
      "loss": 0.6979,
      "step": 161
    },
    {
      "epoch": 2.068085106382979,
      "grad_norm": 0.5376619126851392,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.7996,
      "step": 162
    },
    {
      "epoch": 2.0808510638297872,
      "grad_norm": 0.4976912103481773,
      "learning_rate": 2.5650558779781635e-06,
      "loss": 0.5683,
      "step": 163
    },
    {
      "epoch": 2.0936170212765957,
      "grad_norm": 0.49223792144287865,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.6968,
      "step": 164
    },
    {
      "epoch": 2.106382978723404,
      "grad_norm": 0.5474656248665821,
      "learning_rate": 2.43550361297047e-06,
      "loss": 0.8232,
      "step": 165
    },
    {
      "epoch": 2.119148936170213,
      "grad_norm": 0.4800286329768494,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.6172,
      "step": 166
    },
    {
      "epoch": 2.1319148936170214,
      "grad_norm": 0.574376129911797,
      "learning_rate": 2.3082469195465893e-06,
      "loss": 0.7489,
      "step": 167
    },
    {
      "epoch": 2.1446808510638298,
      "grad_norm": 0.537439553766131,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.6874,
      "step": 168
    },
    {
      "epoch": 2.157446808510638,
      "grad_norm": 0.5254400699435705,
      "learning_rate": 2.1833997096818897e-06,
      "loss": 0.6266,
      "step": 169
    },
    {
      "epoch": 2.1702127659574466,
      "grad_norm": 0.5028076699072636,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.6489,
      "step": 170
    },
    {
      "epoch": 2.1829787234042555,
      "grad_norm": 0.586944387365733,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.763,
      "step": 171
    },
    {
      "epoch": 2.195744680851064,
      "grad_norm": 0.505894086651978,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.6465,
      "step": 172
    },
    {
      "epoch": 2.2085106382978723,
      "grad_norm": 0.53984278729039,
      "learning_rate": 1.941378504424968e-06,
      "loss": 0.6834,
      "step": 173
    },
    {
      "epoch": 2.2212765957446807,
      "grad_norm": 0.5072572058751608,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.7395,
      "step": 174
    },
    {
      "epoch": 2.2340425531914896,
      "grad_norm": 0.515782601163649,
      "learning_rate": 1.8244211507891064e-06,
      "loss": 0.7249,
      "step": 175
    },
    {
      "epoch": 2.246808510638298,
      "grad_norm": 0.496774063368525,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.7169,
      "step": 176
    },
    {
      "epoch": 2.2595744680851064,
      "grad_norm": 0.5168551876746815,
      "learning_rate": 1.7103063703014372e-06,
      "loss": 0.7966,
      "step": 177
    },
    {
      "epoch": 2.272340425531915,
      "grad_norm": 0.5301738065791539,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.6413,
      "step": 178
    },
    {
      "epoch": 2.2851063829787233,
      "grad_norm": 0.5082369172245946,
      "learning_rate": 1.5991363111454023e-06,
      "loss": 0.6976,
      "step": 179
    },
    {
      "epoch": 2.297872340425532,
      "grad_norm": 0.4793427755174214,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.6723,
      "step": 180
    },
    {
      "epoch": 2.3106382978723405,
      "grad_norm": 0.4752348288080126,
      "learning_rate": 1.4910104855800429e-06,
      "loss": 0.7238,
      "step": 181
    },
    {
      "epoch": 2.323404255319149,
      "grad_norm": 0.5266752732027756,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.7487,
      "step": 182
    },
    {
      "epoch": 2.3361702127659574,
      "grad_norm": 0.4469475413548446,
      "learning_rate": 1.3860256808630429e-06,
      "loss": 0.6416,
      "step": 183
    },
    {
      "epoch": 2.348936170212766,
      "grad_norm": 0.4644626840971213,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.7321,
      "step": 184
    },
    {
      "epoch": 2.3617021276595747,
      "grad_norm": 0.508650359767618,
      "learning_rate": 1.2842758726130283e-06,
      "loss": 0.7622,
      "step": 185
    },
    {
      "epoch": 2.374468085106383,
      "grad_norm": 0.5007379233254791,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.6713,
      "step": 186
    },
    {
      "epoch": 2.3872340425531915,
      "grad_norm": 0.49661728721534804,
      "learning_rate": 1.1858521406886674e-06,
      "loss": 0.7533,
      "step": 187
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.5053348466732649,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.6469,
      "step": 188
    },
    {
      "epoch": 2.4127659574468083,
      "grad_norm": 0.5050705783638247,
      "learning_rate": 1.0908425876598512e-06,
      "loss": 0.6918,
      "step": 189
    },
    {
      "epoch": 2.425531914893617,
      "grad_norm": 0.4690909098789046,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.6792,
      "step": 190
    },
    {
      "epoch": 2.4382978723404256,
      "grad_norm": 0.5066403427414923,
      "learning_rate": 9.993322599439692e-07,
      "loss": 0.7389,
      "step": 191
    },
    {
      "epoch": 2.451063829787234,
      "grad_norm": 0.5085674433952515,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.6471,
      "step": 192
    },
    {
      "epoch": 2.4638297872340424,
      "grad_norm": 0.5306625504474757,
      "learning_rate": 9.114030716778433e-07,
      "loss": 0.7413,
      "step": 193
    },
    {
      "epoch": 2.476595744680851,
      "grad_norm": 0.4383722207579253,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.6261,
      "step": 194
    },
    {
      "epoch": 2.4893617021276597,
      "grad_norm": 0.46044975004169747,
      "learning_rate": 8.271337313934869e-07,
      "loss": 0.6867,
      "step": 195
    },
    {
      "epoch": 2.502127659574468,
      "grad_norm": 0.5149550991848054,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.7692,
      "step": 196
    },
    {
      "epoch": 2.5148936170212766,
      "grad_norm": 0.47405804721356093,
      "learning_rate": 7.465996715633028e-07,
      "loss": 0.6737,
      "step": 197
    },
    {
      "epoch": 2.527659574468085,
      "grad_norm": 0.4697866538815745,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.7463,
      "step": 198
    },
    {
      "epoch": 2.5404255319148934,
      "grad_norm": 0.45370547180350884,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.7835,
      "step": 199
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 0.447877218083466,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.7075,
      "step": 200
    },
    {
      "epoch": 2.5659574468085107,
      "grad_norm": 0.4851283761699553,
      "learning_rate": 5.9702234071631e-07,
      "loss": 0.7195,
      "step": 201
    },
    {
      "epoch": 2.578723404255319,
      "grad_norm": 0.4562237608290687,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.6318,
      "step": 202
    },
    {
      "epoch": 2.5914893617021275,
      "grad_norm": 0.5050620555976538,
      "learning_rate": 5.281129616683167e-07,
      "loss": 0.7642,
      "step": 203
    },
    {
      "epoch": 2.604255319148936,
      "grad_norm": 0.49378376497238,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.7023,
      "step": 204
    },
    {
      "epoch": 2.617021276595745,
      "grad_norm": 0.5128163218131276,
      "learning_rate": 4.632065271606756e-07,
      "loss": 0.7142,
      "step": 205
    },
    {
      "epoch": 2.629787234042553,
      "grad_norm": 0.4988207702725477,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.6562,
      "step": 206
    },
    {
      "epoch": 2.6425531914893616,
      "grad_norm": 0.4681720156840467,
      "learning_rate": 4.0236113724274716e-07,
      "loss": 0.7008,
      "step": 207
    },
    {
      "epoch": 2.65531914893617,
      "grad_norm": 0.4892617801756113,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.7654,
      "step": 208
    },
    {
      "epoch": 2.6680851063829785,
      "grad_norm": 0.5131199581603079,
      "learning_rate": 3.4563125677897936e-07,
      "loss": 0.6106,
      "step": 209
    },
    {
      "epoch": 2.6808510638297873,
      "grad_norm": 0.5061855484479584,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.8295,
      "step": 210
    },
    {
      "epoch": 2.6936170212765957,
      "grad_norm": 0.4532598152764269,
      "learning_rate": 2.930676666954846e-07,
      "loss": 0.6481,
      "step": 211
    },
    {
      "epoch": 2.706382978723404,
      "grad_norm": 0.46463132032117166,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.642,
      "step": 212
    },
    {
      "epoch": 2.719148936170213,
      "grad_norm": 0.49012449300116195,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.7463,
      "step": 213
    },
    {
      "epoch": 2.731914893617021,
      "grad_norm": 0.4636477810079059,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.6347,
      "step": 214
    },
    {
      "epoch": 2.74468085106383,
      "grad_norm": 0.47971257641576737,
      "learning_rate": 2.006237922855553e-07,
      "loss": 0.8141,
      "step": 215
    },
    {
      "epoch": 2.7574468085106383,
      "grad_norm": 0.4522928987229415,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.6098,
      "step": 216
    },
    {
      "epoch": 2.7702127659574467,
      "grad_norm": 0.4708467076318233,
      "learning_rate": 1.6082625774666793e-07,
      "loss": 0.7974,
      "step": 217
    },
    {
      "epoch": 2.7829787234042556,
      "grad_norm": 0.45260252635853443,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.6051,
      "step": 218
    },
    {
      "epoch": 2.795744680851064,
      "grad_norm": 0.4772874712122134,
      "learning_rate": 1.253604390908819e-07,
      "loss": 0.7594,
      "step": 219
    },
    {
      "epoch": 2.8085106382978724,
      "grad_norm": 0.4379365885870637,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.6206,
      "step": 220
    },
    {
      "epoch": 2.821276595744681,
      "grad_norm": 0.5353989006280159,
      "learning_rate": 9.42580830291373e-08,
      "loss": 0.8591,
      "step": 221
    },
    {
      "epoch": 2.8340425531914892,
      "grad_norm": 0.4388978526940481,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.6012,
      "step": 222
    },
    {
      "epoch": 2.846808510638298,
      "grad_norm": 0.4627380722362225,
      "learning_rate": 6.75470303823933e-08,
      "loss": 0.6869,
      "step": 223
    },
    {
      "epoch": 2.8595744680851065,
      "grad_norm": 0.5100171595604029,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.6985,
      "step": 224
    },
    {
      "epoch": 2.872340425531915,
      "grad_norm": 0.5163263210032598,
      "learning_rate": 4.52511911603265e-08,
      "loss": 0.7723,
      "step": 225
    },
    {
      "epoch": 2.8851063829787233,
      "grad_norm": 0.4703957778516729,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.6751,
      "step": 226
    },
    {
      "epoch": 2.8978723404255318,
      "grad_norm": 0.4353996932568037,
      "learning_rate": 2.7390523158633552e-08,
      "loss": 0.6796,
      "step": 227
    },
    {
      "epoch": 2.9106382978723406,
      "grad_norm": 0.5329335872850072,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.7351,
      "step": 228
    },
    {
      "epoch": 2.923404255319149,
      "grad_norm": 0.5187235806914524,
      "learning_rate": 1.3981014094099354e-08,
      "loss": 0.6504,
      "step": 229
    },
    {
      "epoch": 2.9361702127659575,
      "grad_norm": 0.49115304636235096,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.6921,
      "step": 230
    },
    {
      "epoch": 2.948936170212766,
      "grad_norm": 0.5169642162297178,
      "learning_rate": 5.034667293427053e-09,
      "loss": 0.7357,
      "step": 231
    },
    {
      "epoch": 2.9617021276595743,
      "grad_norm": 0.467272931984639,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.6568,
      "step": 232
    },
    {
      "epoch": 2.974468085106383,
      "grad_norm": 0.47352407416977127,
      "learning_rate": 5.594909486328348e-10,
      "loss": 0.6961,
      "step": 233
    },
    {
      "epoch": 2.9872340425531916,
      "grad_norm": 0.46921973983708287,
      "learning_rate": 0.0,
      "loss": 0.6174,
      "step": 234
    },
    {
      "epoch": 2.9872340425531916,
      "step": 234,
      "total_flos": 182640614014976.0,
      "train_loss": 0.7958560710788792,
      "train_runtime": 3531.6623,
      "train_samples_per_second": 6.371,
      "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": 182640614014976.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}