{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.968152866242038,
  "eval_steps": 500,
  "global_step": 234,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012738853503184714,
      "grad_norm": 8.287943840026855,
      "learning_rate": 0.0,
      "loss": 2.1051,
      "mean_token_accuracy": 0.6576112508773804,
      "num_tokens": 7722.0,
      "step": 1
    },
    {
      "epoch": 0.025477707006369428,
      "grad_norm": 8.355767250061035,
      "learning_rate": 2.5e-05,
      "loss": 2.0935,
      "mean_token_accuracy": 0.662928581237793,
      "num_tokens": 15318.0,
      "step": 2
    },
    {
      "epoch": 0.03821656050955414,
      "grad_norm": 5.267609119415283,
      "learning_rate": 5e-05,
      "loss": 1.823,
      "mean_token_accuracy": 0.6904586851596832,
      "num_tokens": 23143.0,
      "step": 3
    },
    {
      "epoch": 0.050955414012738856,
      "grad_norm": 4.595154762268066,
      "learning_rate": 7.500000000000001e-05,
      "loss": 1.3159,
      "mean_token_accuracy": 0.7333050668239594,
      "num_tokens": 30786.0,
      "step": 4
    },
    {
      "epoch": 0.06369426751592357,
      "grad_norm": 1.5862985849380493,
      "learning_rate": 0.0001,
      "loss": 1.0104,
      "mean_token_accuracy": 0.7833340167999268,
      "num_tokens": 38345.0,
      "step": 5
    },
    {
      "epoch": 0.07643312101910828,
      "grad_norm": 0.8021528124809265,
      "learning_rate": 0.000125,
      "loss": 0.8478,
      "mean_token_accuracy": 0.8248030543327332,
      "num_tokens": 45956.0,
      "step": 6
    },
    {
      "epoch": 0.08917197452229299,
      "grad_norm": 0.5970975756645203,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.7815,
      "mean_token_accuracy": 0.8295383155345917,
      "num_tokens": 53562.0,
      "step": 7
    },
    {
      "epoch": 0.10191082802547771,
      "grad_norm": 1.414368987083435,
      "learning_rate": 0.000175,
      "loss": 0.7247,
      "mean_token_accuracy": 0.8384078443050385,
      "num_tokens": 60929.0,
      "step": 8
    },
    {
      "epoch": 0.11464968152866242,
      "grad_norm": 0.9989311099052429,
      "learning_rate": 0.0002,
      "loss": 0.74,
      "mean_token_accuracy": 0.8326991200447083,
      "num_tokens": 68327.0,
      "step": 9
    },
    {
      "epoch": 0.12738853503184713,
      "grad_norm": 0.4296531081199646,
      "learning_rate": 0.00019999033847063811,
      "loss": 0.7706,
      "mean_token_accuracy": 0.8299016058444977,
      "num_tokens": 75754.0,
      "step": 10
    },
    {
      "epoch": 0.14012738853503184,
      "grad_norm": 0.4344661831855774,
      "learning_rate": 0.00019996135574945544,
      "loss": 0.7121,
      "mean_token_accuracy": 0.8349876701831818,
      "num_tokens": 83217.0,
      "step": 11
    },
    {
      "epoch": 0.15286624203821655,
      "grad_norm": 0.3851865530014038,
      "learning_rate": 0.00019991305743680013,
      "loss": 0.747,
      "mean_token_accuracy": 0.830773264169693,
      "num_tokens": 90805.0,
      "step": 12
    },
    {
      "epoch": 0.16560509554140126,
      "grad_norm": 0.3876875042915344,
      "learning_rate": 0.0001998454528653836,
      "loss": 0.7502,
      "mean_token_accuracy": 0.8285396993160248,
      "num_tokens": 98404.0,
      "step": 13
    },
    {
      "epoch": 0.17834394904458598,
      "grad_norm": 0.4084186553955078,
      "learning_rate": 0.00019975855509847686,
      "loss": 0.7135,
      "mean_token_accuracy": 0.83475062251091,
      "num_tokens": 106081.0,
      "step": 14
    },
    {
      "epoch": 0.1910828025477707,
      "grad_norm": 0.3862064480781555,
      "learning_rate": 0.00019965238092738643,
      "loss": 0.777,
      "mean_token_accuracy": 0.8248996436595917,
      "num_tokens": 113617.0,
      "step": 15
    },
    {
      "epoch": 0.20382165605095542,
      "grad_norm": 0.4174270033836365,
      "learning_rate": 0.00019952695086820975,
      "loss": 0.7301,
      "mean_token_accuracy": 0.8283782601356506,
      "num_tokens": 121314.0,
      "step": 16
    },
    {
      "epoch": 0.21656050955414013,
      "grad_norm": 0.3616231381893158,
      "learning_rate": 0.0001993822891578708,
      "loss": 0.7061,
      "mean_token_accuracy": 0.8416012227535248,
      "num_tokens": 128659.0,
      "step": 17
    },
    {
      "epoch": 0.22929936305732485,
      "grad_norm": 0.38920992612838745,
      "learning_rate": 0.0001992184237494368,
      "loss": 0.7001,
      "mean_token_accuracy": 0.835535317659378,
      "num_tokens": 136314.0,
      "step": 18
    },
    {
      "epoch": 0.24203821656050956,
      "grad_norm": 0.3504973351955414,
      "learning_rate": 0.0001990353863067169,
      "loss": 0.7106,
      "mean_token_accuracy": 0.8328219652175903,
      "num_tokens": 144182.0,
      "step": 19
    },
    {
      "epoch": 0.25477707006369427,
      "grad_norm": 0.34746506810188293,
      "learning_rate": 0.0001988332121981436,
      "loss": 0.7584,
      "mean_token_accuracy": 0.8270872235298157,
      "num_tokens": 152095.0,
      "step": 20
    },
    {
      "epoch": 0.267515923566879,
      "grad_norm": 0.3385103642940521,
      "learning_rate": 0.00019861194048993863,
      "loss": 0.6867,
      "mean_token_accuracy": 0.8429812490940094,
      "num_tokens": 159463.0,
      "step": 21
    },
    {
      "epoch": 0.2802547770700637,
      "grad_norm": 0.3274126946926117,
      "learning_rate": 0.0001983716139385641,
      "loss": 0.6727,
      "mean_token_accuracy": 0.8409291505813599,
      "num_tokens": 166878.0,
      "step": 22
    },
    {
      "epoch": 0.2929936305732484,
      "grad_norm": 0.33483415842056274,
      "learning_rate": 0.0001981122789824607,
      "loss": 0.656,
      "mean_token_accuracy": 0.8487405478954315,
      "num_tokens": 174314.0,
      "step": 23
    },
    {
      "epoch": 0.3057324840764331,
      "grad_norm": 0.3158458173274994,
      "learning_rate": 0.00019783398573307428,
      "loss": 0.7133,
      "mean_token_accuracy": 0.8301410675048828,
      "num_tokens": 181966.0,
      "step": 24
    },
    {
      "epoch": 0.3184713375796178,
      "grad_norm": 0.3395407199859619,
      "learning_rate": 0.00019753678796517282,
      "loss": 0.7214,
      "mean_token_accuracy": 0.8315837681293488,
      "num_tokens": 189637.0,
      "step": 25
    },
    {
      "epoch": 0.33121019108280253,
      "grad_norm": 0.3301332890987396,
      "learning_rate": 0.00019722074310645553,
      "loss": 0.65,
      "mean_token_accuracy": 0.8509158194065094,
      "num_tokens": 196866.0,
      "step": 26
    },
    {
      "epoch": 0.34394904458598724,
      "grad_norm": 0.40064746141433716,
      "learning_rate": 0.00019688591222645607,
      "loss": 0.6879,
      "mean_token_accuracy": 0.8388471305370331,
      "num_tokens": 204527.0,
      "step": 27
    },
    {
      "epoch": 0.35668789808917195,
      "grad_norm": 0.36539191007614136,
      "learning_rate": 0.000196532360024742,
      "loss": 0.7129,
      "mean_token_accuracy": 0.8327633142471313,
      "num_tokens": 212217.0,
      "step": 28
    },
    {
      "epoch": 0.36942675159235666,
      "grad_norm": 0.32001280784606934,
      "learning_rate": 0.0001961601548184129,
      "loss": 0.6097,
      "mean_token_accuracy": 0.8550100326538086,
      "num_tokens": 219611.0,
      "step": 29
    },
    {
      "epoch": 0.3821656050955414,
      "grad_norm": 0.3553815484046936,
      "learning_rate": 0.00019576936852889936,
      "loss": 0.664,
      "mean_token_accuracy": 0.8454612195491791,
      "num_tokens": 227201.0,
      "step": 30
    },
    {
      "epoch": 0.39490445859872614,
      "grad_norm": 0.37485066056251526,
      "learning_rate": 0.00019536007666806556,
      "loss": 0.6167,
      "mean_token_accuracy": 0.8543617725372314,
      "num_tokens": 234596.0,
      "step": 31
    },
    {
      "epoch": 0.40764331210191085,
      "grad_norm": 0.43772801756858826,
      "learning_rate": 0.0001949323583236181,
      "loss": 0.6297,
      "mean_token_accuracy": 0.848662257194519,
      "num_tokens": 242008.0,
      "step": 32
    },
    {
      "epoch": 0.42038216560509556,
      "grad_norm": 0.4182124137878418,
      "learning_rate": 0.0001944862961438239,
      "loss": 0.6675,
      "mean_token_accuracy": 0.8380049467086792,
      "num_tokens": 249776.0,
      "step": 33
    },
    {
      "epoch": 0.43312101910828027,
      "grad_norm": 0.3301289975643158,
      "learning_rate": 0.00019402197632153992,
      "loss": 0.6575,
      "mean_token_accuracy": 0.8471685945987701,
      "num_tokens": 257319.0,
      "step": 34
    },
    {
      "epoch": 0.445859872611465,
      "grad_norm": 0.35300567746162415,
      "learning_rate": 0.00019353948857755803,
      "loss": 0.5979,
      "mean_token_accuracy": 0.8560113310813904,
      "num_tokens": 264828.0,
      "step": 35
    },
    {
      "epoch": 0.4585987261146497,
      "grad_norm": 0.3664567768573761,
      "learning_rate": 0.00019303892614326836,
      "loss": 0.5948,
      "mean_token_accuracy": 0.8638049066066742,
      "num_tokens": 272172.0,
      "step": 36
    },
    {
      "epoch": 0.4713375796178344,
      "grad_norm": 0.3352532386779785,
      "learning_rate": 0.00019252038574264405,
      "loss": 0.6585,
      "mean_token_accuracy": 0.844391405582428,
      "num_tokens": 279813.0,
      "step": 37
    },
    {
      "epoch": 0.4840764331210191,
      "grad_norm": 0.4468683898448944,
      "learning_rate": 0.00019198396757355118,
      "loss": 0.6146,
      "mean_token_accuracy": 0.8511447310447693,
      "num_tokens": 287537.0,
      "step": 38
    },
    {
      "epoch": 0.4968152866242038,
      "grad_norm": 0.3902146518230438,
      "learning_rate": 0.00019142977528838762,
      "loss": 0.5993,
      "mean_token_accuracy": 0.8533865213394165,
      "num_tokens": 295085.0,
      "step": 39
    },
    {
      "epoch": 0.5095541401273885,
      "grad_norm": 0.3714890778064728,
      "learning_rate": 0.00019085791597405404,
      "loss": 0.6601,
      "mean_token_accuracy": 0.8435452580451965,
      "num_tokens": 302853.0,
      "step": 40
    },
    {
      "epoch": 0.5222929936305732,
      "grad_norm": 0.35523873567581177,
      "learning_rate": 0.00019026850013126157,
      "loss": 0.6153,
      "mean_token_accuracy": 0.8525647222995758,
      "num_tokens": 310507.0,
      "step": 41
    },
    {
      "epoch": 0.535031847133758,
      "grad_norm": 0.35029295086860657,
      "learning_rate": 0.00018966164165317966,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8702357113361359,
      "num_tokens": 317912.0,
      "step": 42
    },
    {
      "epoch": 0.5477707006369427,
      "grad_norm": 0.38241031765937805,
      "learning_rate": 0.00018903745780342839,
      "loss": 0.6103,
      "mean_token_accuracy": 0.855758547782898,
      "num_tokens": 325400.0,
      "step": 43
    },
    {
      "epoch": 0.5605095541401274,
      "grad_norm": 0.4096700847148895,
      "learning_rate": 0.0001883960691934196,
      "loss": 0.5746,
      "mean_token_accuracy": 0.8615990579128265,
      "num_tokens": 333086.0,
      "step": 44
    },
    {
      "epoch": 0.5732484076433121,
      "grad_norm": 0.37658724188804626,
      "learning_rate": 0.00018773759975905098,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8693232238292694,
      "num_tokens": 340578.0,
      "step": 45
    },
    {
      "epoch": 0.5859872611464968,
      "grad_norm": 0.43547865748405457,
      "learning_rate": 0.00018706217673675811,
      "loss": 0.6485,
      "mean_token_accuracy": 0.8440962433815002,
      "num_tokens": 348586.0,
      "step": 46
    },
    {
      "epoch": 0.5987261146496815,
      "grad_norm": 0.4606596827507019,
      "learning_rate": 0.0001863699306389282,
      "loss": 0.5552,
      "mean_token_accuracy": 0.8659605085849762,
      "num_tokens": 356174.0,
      "step": 47
    },
    {
      "epoch": 0.6114649681528662,
      "grad_norm": 0.4326448142528534,
      "learning_rate": 0.00018566099522868119,
      "loss": 0.5145,
      "mean_token_accuracy": 0.8702480494976044,
      "num_tokens": 363805.0,
      "step": 48
    },
    {
      "epoch": 0.6242038216560509,
      "grad_norm": 0.4241434633731842,
      "learning_rate": 0.00018493550749402278,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8582789897918701,
      "num_tokens": 371462.0,
      "step": 49
    },
    {
      "epoch": 0.6369426751592356,
      "grad_norm": 0.4271712005138397,
      "learning_rate": 0.00018419360762137395,
      "loss": 0.5618,
      "mean_token_accuracy": 0.8646855652332306,
      "num_tokens": 379019.0,
      "step": 50
    },
    {
      "epoch": 0.6496815286624203,
      "grad_norm": 0.40694111585617065,
      "learning_rate": 0.00018343543896848273,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8710185289382935,
      "num_tokens": 386380.0,
      "step": 51
    },
    {
      "epoch": 0.6624203821656051,
      "grad_norm": 0.40776747465133667,
      "learning_rate": 0.00018266114803672318,
      "loss": 0.5566,
      "mean_token_accuracy": 0.8674687147140503,
      "num_tokens": 393867.0,
      "step": 52
    },
    {
      "epoch": 0.6751592356687898,
      "grad_norm": 0.4121972620487213,
      "learning_rate": 0.00018187088444278674,
      "loss": 0.5482,
      "mean_token_accuracy": 0.8711592853069305,
      "num_tokens": 401609.0,
      "step": 53
    },
    {
      "epoch": 0.6878980891719745,
      "grad_norm": 0.40399909019470215,
      "learning_rate": 0.00018106480088977172,
      "loss": 0.5687,
      "mean_token_accuracy": 0.8622486889362335,
      "num_tokens": 409264.0,
      "step": 54
    },
    {
      "epoch": 0.7006369426751592,
      "grad_norm": 0.41347479820251465,
      "learning_rate": 0.00018024305313767646,
      "loss": 0.5048,
      "mean_token_accuracy": 0.8774301111698151,
      "num_tokens": 416527.0,
      "step": 55
    },
    {
      "epoch": 0.7133757961783439,
      "grad_norm": 0.44538500905036926,
      "learning_rate": 0.00017940579997330165,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8834721446037292,
      "num_tokens": 424091.0,
      "step": 56
    },
    {
      "epoch": 0.7261146496815286,
      "grad_norm": 0.5161186456680298,
      "learning_rate": 0.00017855320317956784,
      "loss": 0.5374,
      "mean_token_accuracy": 0.867156058549881,
      "num_tokens": 431859.0,
      "step": 57
    },
    {
      "epoch": 0.7388535031847133,
      "grad_norm": 0.4894382953643799,
      "learning_rate": 0.00017768542750425426,
      "loss": 0.5479,
      "mean_token_accuracy": 0.8681759834289551,
      "num_tokens": 439429.0,
      "step": 58
    },
    {
      "epoch": 0.7515923566878981,
      "grad_norm": 0.4625523090362549,
      "learning_rate": 0.0001768026406281642,
      "loss": 0.587,
      "mean_token_accuracy": 0.8580531179904938,
      "num_tokens": 447417.0,
      "step": 59
    },
    {
      "epoch": 0.7643312101910829,
      "grad_norm": 0.4163782000541687,
      "learning_rate": 0.00017590501313272415,
      "loss": 0.5592,
      "mean_token_accuracy": 0.868916928768158,
      "num_tokens": 455053.0,
      "step": 60
    },
    {
      "epoch": 0.7770700636942676,
      "grad_norm": 0.4515700042247772,
      "learning_rate": 0.00017499271846702213,
      "loss": 0.4692,
      "mean_token_accuracy": 0.8839752674102783,
      "num_tokens": 462553.0,
      "step": 61
    },
    {
      "epoch": 0.7898089171974523,
      "grad_norm": 0.48304829001426697,
      "learning_rate": 0.00017406593291429217,
      "loss": 0.4951,
      "mean_token_accuracy": 0.8761360943317413,
      "num_tokens": 470057.0,
      "step": 62
    },
    {
      "epoch": 0.802547770700637,
      "grad_norm": 0.4670911729335785,
      "learning_rate": 0.00017312483555785086,
      "loss": 0.4412,
      "mean_token_accuracy": 0.8923192024230957,
      "num_tokens": 477410.0,
      "step": 63
    },
    {
      "epoch": 0.8152866242038217,
      "grad_norm": 0.4475160241127014,
      "learning_rate": 0.00017216960824649303,
      "loss": 0.4687,
      "mean_token_accuracy": 0.8837877810001373,
      "num_tokens": 485000.0,
      "step": 64
    },
    {
      "epoch": 0.8280254777070064,
      "grad_norm": 0.4836560785770416,
      "learning_rate": 0.00017120043555935298,
      "loss": 0.4642,
      "mean_token_accuracy": 0.886978417634964,
      "num_tokens": 492576.0,
      "step": 65
    },
    {
      "epoch": 0.8407643312101911,
      "grad_norm": 0.5065037608146667,
      "learning_rate": 0.0001702175047702382,
      "loss": 0.5121,
      "mean_token_accuracy": 0.8710624575614929,
      "num_tokens": 500413.0,
      "step": 66
    },
    {
      "epoch": 0.8535031847133758,
      "grad_norm": 0.4843185245990753,
      "learning_rate": 0.00016922100581144228,
      "loss": 0.5384,
      "mean_token_accuracy": 0.8697362542152405,
      "num_tokens": 508117.0,
      "step": 67
    },
    {
      "epoch": 0.8662420382165605,
      "grad_norm": 0.5202386379241943,
      "learning_rate": 0.00016821113123704424,
      "loss": 0.4639,
      "mean_token_accuracy": 0.8836866319179535,
      "num_tokens": 515746.0,
      "step": 68
    },
    {
      "epoch": 0.8789808917197452,
      "grad_norm": 0.4895678460597992,
      "learning_rate": 0.00016718807618570106,
      "loss": 0.4422,
      "mean_token_accuracy": 0.8906868398189545,
      "num_tokens": 523013.0,
      "step": 69
    },
    {
      "epoch": 0.89171974522293,
      "grad_norm": 0.5349318981170654,
      "learning_rate": 0.00016615203834294119,
      "loss": 0.4428,
      "mean_token_accuracy": 0.8886107206344604,
      "num_tokens": 530452.0,
      "step": 70
    },
    {
      "epoch": 0.9044585987261147,
      "grad_norm": 0.48844024538993835,
      "learning_rate": 0.00016510321790296525,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8925909399986267,
      "num_tokens": 537855.0,
      "step": 71
    },
    {
      "epoch": 0.9171974522292994,
      "grad_norm": 0.5755529999732971,
      "learning_rate": 0.00016404181752996289,
      "loss": 0.4864,
      "mean_token_accuracy": 0.8765636086463928,
      "num_tokens": 545421.0,
      "step": 72
    },
    {
      "epoch": 0.9299363057324841,
      "grad_norm": 0.45479002594947815,
      "learning_rate": 0.00016296804231895142,
      "loss": 0.3651,
      "mean_token_accuracy": 0.9048753976821899,
      "num_tokens": 552909.0,
      "step": 73
    },
    {
      "epoch": 0.9426751592356688,
      "grad_norm": 0.49108317494392395,
      "learning_rate": 0.00016188209975614542,
      "loss": 0.45,
      "mean_token_accuracy": 0.8858565986156464,
      "num_tokens": 560702.0,
      "step": 74
    },
    {
      "epoch": 0.9554140127388535,
      "grad_norm": 0.46213510632514954,
      "learning_rate": 0.00016078419967886402,
      "loss": 0.3958,
      "mean_token_accuracy": 0.9005838930606842,
      "num_tokens": 568245.0,
      "step": 75
    },
    {
      "epoch": 0.9681528662420382,
      "grad_norm": 0.5798938870429993,
      "learning_rate": 0.00015967455423498387,
      "loss": 0.3912,
      "mean_token_accuracy": 0.9034214019775391,
      "num_tokens": 575857.0,
      "step": 76
    },
    {
      "epoch": 0.9808917197452229,
      "grad_norm": 0.5185940861701965,
      "learning_rate": 0.00015855337784194577,
      "loss": 0.3628,
      "mean_token_accuracy": 0.9065321683883667,
      "num_tokens": 583441.0,
      "step": 77
    },
    {
      "epoch": 0.9936305732484076,
      "grad_norm": 0.5182363390922546,
      "learning_rate": 0.00015742088714532247,
      "loss": 0.4174,
      "mean_token_accuracy": 0.8943608999252319,
      "num_tokens": 591106.0,
      "step": 78
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5182363390922546,
      "learning_rate": 0.00015627730097695638,
      "loss": 0.4455,
      "mean_token_accuracy": 0.8872870206832886,
      "num_tokens": 594953.0,
      "step": 79
    },
    {
      "epoch": 1.0127388535031847,
      "grad_norm": 0.823184072971344,
      "learning_rate": 0.00015512284031267437,
      "loss": 0.2927,
      "mean_token_accuracy": 0.9254479706287384,
      "num_tokens": 602519.0,
      "step": 80
    },
    {
      "epoch": 1.0254777070063694,
      "grad_norm": 0.4461751878261566,
      "learning_rate": 0.00015395772822958845,
      "loss": 0.2936,
      "mean_token_accuracy": 0.9253770411014557,
      "num_tokens": 610085.0,
      "step": 81
    },
    {
      "epoch": 1.0382165605095541,
      "grad_norm": 0.5058058500289917,
      "learning_rate": 0.00015278218986299074,
      "loss": 0.235,
      "mean_token_accuracy": 0.9344533979892731,
      "num_tokens": 617710.0,
      "step": 82
    },
    {
      "epoch": 1.0509554140127388,
      "grad_norm": 0.6525824069976807,
      "learning_rate": 0.0001515964523628501,
      "loss": 0.2901,
      "mean_token_accuracy": 0.9262363314628601,
      "num_tokens": 625223.0,
      "step": 83
    },
    {
      "epoch": 1.0636942675159236,
      "grad_norm": 0.6204919815063477,
      "learning_rate": 0.00015040074484992,
      "loss": 0.2371,
      "mean_token_accuracy": 0.9375110268592834,
      "num_tokens": 632672.0,
      "step": 84
    },
    {
      "epoch": 1.0764331210191083,
      "grad_norm": 0.4697447419166565,
      "learning_rate": 0.00014919529837146528,
      "loss": 0.2389,
      "mean_token_accuracy": 0.9337205290794373,
      "num_tokens": 640267.0,
      "step": 85
    },
    {
      "epoch": 1.089171974522293,
      "grad_norm": 0.5069385170936584,
      "learning_rate": 0.00014798034585661695,
      "loss": 0.2394,
      "mean_token_accuracy": 0.9344863593578339,
      "num_tokens": 647855.0,
      "step": 86
    },
    {
      "epoch": 1.1019108280254777,
      "grad_norm": 0.512169599533081,
      "learning_rate": 0.0001467561220713628,
      "loss": 0.3004,
      "mean_token_accuracy": 0.9232901930809021,
      "num_tokens": 655347.0,
      "step": 87
    },
    {
      "epoch": 1.1146496815286624,
      "grad_norm": 0.5167611837387085,
      "learning_rate": 0.0001455228635731839,
      "loss": 0.261,
      "mean_token_accuracy": 0.9302393794059753,
      "num_tokens": 662908.0,
      "step": 88
    },
    {
      "epoch": 1.127388535031847,
      "grad_norm": 0.43430718779563904,
      "learning_rate": 0.00014428080866534396,
      "loss": 0.2426,
      "mean_token_accuracy": 0.9376973807811737,
      "num_tokens": 670227.0,
      "step": 89
    },
    {
      "epoch": 1.1401273885350318,
      "grad_norm": 0.44046974182128906,
      "learning_rate": 0.00014303019735084226,
      "loss": 0.2484,
      "mean_token_accuracy": 0.9353660345077515,
      "num_tokens": 677733.0,
      "step": 90
    },
    {
      "epoch": 1.1528662420382165,
      "grad_norm": 0.5374375581741333,
      "learning_rate": 0.00014177127128603745,
      "loss": 0.2681,
      "mean_token_accuracy": 0.9323221743106842,
      "num_tokens": 685450.0,
      "step": 91
    },
    {
      "epoch": 1.1656050955414012,
      "grad_norm": 0.507544219493866,
      "learning_rate": 0.0001405042737339524,
      "loss": 0.2522,
      "mean_token_accuracy": 0.9334528148174286,
      "num_tokens": 693379.0,
      "step": 92
    },
    {
      "epoch": 1.178343949044586,
      "grad_norm": 0.5681467056274414,
      "learning_rate": 0.0001392294495172681,
      "loss": 0.2381,
      "mean_token_accuracy": 0.9384985864162445,
      "num_tokens": 701006.0,
      "step": 93
    },
    {
      "epoch": 1.1910828025477707,
      "grad_norm": 0.5037748217582703,
      "learning_rate": 0.00013794704497101655,
      "loss": 0.248,
      "mean_token_accuracy": 0.9311998784542084,
      "num_tokens": 708222.0,
      "step": 94
    },
    {
      "epoch": 1.2038216560509554,
      "grad_norm": 0.4543361961841583,
      "learning_rate": 0.0001366573078949813,
      "loss": 0.2709,
      "mean_token_accuracy": 0.927354484796524,
      "num_tokens": 715815.0,
      "step": 95
    },
    {
      "epoch": 1.21656050955414,
      "grad_norm": 0.47324758768081665,
      "learning_rate": 0.00013536048750581494,
      "loss": 0.2349,
      "mean_token_accuracy": 0.941594272851944,
      "num_tokens": 723177.0,
      "step": 96
    },
    {
      "epoch": 1.2292993630573248,
      "grad_norm": 0.4437101185321808,
      "learning_rate": 0.00013405683438888282,
      "loss": 0.2264,
      "mean_token_accuracy": 0.9379465878009796,
      "num_tokens": 730685.0,
      "step": 97
    },
    {
      "epoch": 1.2420382165605095,
      "grad_norm": 0.3893497884273529,
      "learning_rate": 0.00013274660044984224,
      "loss": 0.2139,
      "mean_token_accuracy": 0.9428676962852478,
      "num_tokens": 738399.0,
      "step": 98
    },
    {
      "epoch": 1.2547770700636942,
      "grad_norm": 0.47590160369873047,
      "learning_rate": 0.00013143003886596669,
      "loss": 0.2622,
      "mean_token_accuracy": 0.9279411733150482,
      "num_tokens": 746096.0,
      "step": 99
    },
    {
      "epoch": 1.267515923566879,
      "grad_norm": 0.5140976309776306,
      "learning_rate": 0.0001301074040372242,
      "loss": 0.2063,
      "mean_token_accuracy": 0.9456835091114044,
      "num_tokens": 753562.0,
      "step": 100
    },
    {
      "epoch": 1.2802547770700636,
      "grad_norm": 0.438860148191452,
      "learning_rate": 0.00012877895153711935,
      "loss": 0.2382,
      "mean_token_accuracy": 0.9374182820320129,
      "num_tokens": 761282.0,
      "step": 101
    },
    {
      "epoch": 1.2929936305732483,
      "grad_norm": 0.6094561815261841,
      "learning_rate": 0.0001274449380633089,
      "loss": 0.2339,
      "mean_token_accuracy": 0.9405188262462616,
      "num_tokens": 768815.0,
      "step": 102
    },
    {
      "epoch": 1.305732484076433,
      "grad_norm": 0.5217446088790894,
      "learning_rate": 0.00012610562138799978,
      "loss": 0.2705,
      "mean_token_accuracy": 0.9315881133079529,
      "num_tokens": 776490.0,
      "step": 103
    },
    {
      "epoch": 1.3184713375796178,
      "grad_norm": 0.5095704197883606,
      "learning_rate": 0.00012476126030813963,
      "loss": 0.2185,
      "mean_token_accuracy": 0.9398861229419708,
      "num_tokens": 784055.0,
      "step": 104
    },
    {
      "epoch": 1.3312101910828025,
      "grad_norm": 0.5053628087043762,
      "learning_rate": 0.0001234121145954094,
      "loss": 0.2421,
      "mean_token_accuracy": 0.9382818043231964,
      "num_tokens": 791644.0,
      "step": 105
    },
    {
      "epoch": 1.3439490445859872,
      "grad_norm": 0.43147337436676025,
      "learning_rate": 0.0001220584449460274,
      "loss": 0.207,
      "mean_token_accuracy": 0.9434195756912231,
      "num_tokens": 799207.0,
      "step": 106
    },
    {
      "epoch": 1.356687898089172,
      "grad_norm": 0.40620630979537964,
      "learning_rate": 0.00012070051293037492,
      "loss": 0.2302,
      "mean_token_accuracy": 0.9408184885978699,
      "num_tokens": 806854.0,
      "step": 107
    },
    {
      "epoch": 1.3694267515923566,
      "grad_norm": 0.4834441542625427,
      "learning_rate": 0.00011933858094245281,
      "loss": 0.2031,
      "mean_token_accuracy": 0.9470761716365814,
      "num_tokens": 814335.0,
      "step": 108
    },
    {
      "epoch": 1.3821656050955413,
      "grad_norm": 0.45893028378486633,
      "learning_rate": 0.00011797291214917881,
      "loss": 0.2137,
      "mean_token_accuracy": 0.9417379796504974,
      "num_tokens": 822129.0,
      "step": 109
    },
    {
      "epoch": 1.394904458598726,
      "grad_norm": 0.5153368711471558,
      "learning_rate": 0.00011660377043953588,
      "loss": 0.2402,
      "mean_token_accuracy": 0.938307374715805,
      "num_tokens": 830009.0,
      "step": 110
    },
    {
      "epoch": 1.4076433121019107,
      "grad_norm": 0.4056824743747711,
      "learning_rate": 0.0001152314203735805,
      "loss": 0.2025,
      "mean_token_accuracy": 0.9457784295082092,
      "num_tokens": 837339.0,
      "step": 111
    },
    {
      "epoch": 1.4203821656050954,
      "grad_norm": 0.4811604619026184,
      "learning_rate": 0.0001138561271313219,
      "loss": 0.2136,
      "mean_token_accuracy": 0.9439316689968109,
      "num_tokens": 844962.0,
      "step": 112
    },
    {
      "epoch": 1.4331210191082802,
      "grad_norm": 0.4947657585144043,
      "learning_rate": 0.00011247815646148087,
      "loss": 0.228,
      "mean_token_accuracy": 0.9374533593654633,
      "num_tokens": 852442.0,
      "step": 113
    },
    {
      "epoch": 1.4458598726114649,
      "grad_norm": 0.4831300973892212,
      "learning_rate": 0.00011109777463013915,
      "loss": 0.1971,
      "mean_token_accuracy": 0.950575202703476,
      "num_tokens": 859705.0,
      "step": 114
    },
    {
      "epoch": 1.4585987261146496,
      "grad_norm": 0.4768920838832855,
      "learning_rate": 0.0001097152483692886,
      "loss": 0.2014,
      "mean_token_accuracy": 0.9461125135421753,
      "num_tokens": 867320.0,
      "step": 115
    },
    {
      "epoch": 1.4713375796178343,
      "grad_norm": 0.4603787362575531,
      "learning_rate": 0.00010833084482529048,
      "loss": 0.2343,
      "mean_token_accuracy": 0.9413523375988007,
      "num_tokens": 874840.0,
      "step": 116
    },
    {
      "epoch": 1.484076433121019,
      "grad_norm": 0.4399753510951996,
      "learning_rate": 0.00010694483150725458,
      "loss": 0.2075,
      "mean_token_accuracy": 0.948328822851181,
      "num_tokens": 882451.0,
      "step": 117
    },
    {
      "epoch": 1.4968152866242037,
      "grad_norm": 0.483257919549942,
      "learning_rate": 0.00010555747623534831,
      "loss": 0.1985,
      "mean_token_accuracy": 0.9503215551376343,
      "num_tokens": 890101.0,
      "step": 118
    },
    {
      "epoch": 1.5095541401273884,
      "grad_norm": 0.45524635910987854,
      "learning_rate": 0.00010416904708904548,
      "loss": 0.1954,
      "mean_token_accuracy": 0.9504112601280212,
      "num_tokens": 897607.0,
      "step": 119
    },
    {
      "epoch": 1.5222929936305731,
      "grad_norm": 0.44008612632751465,
      "learning_rate": 0.00010277981235532541,
      "loss": 0.212,
      "mean_token_accuracy": 0.9436098635196686,
      "num_tokens": 905191.0,
      "step": 120
    },
    {
      "epoch": 1.5350318471337578,
      "grad_norm": 0.46399325132369995,
      "learning_rate": 0.00010139004047683151,
      "loss": 0.1785,
      "mean_token_accuracy": 0.9537560045719147,
      "num_tokens": 912523.0,
      "step": 121
    },
    {
      "epoch": 1.5477707006369426,
      "grad_norm": 0.4352799654006958,
      "learning_rate": 0.0001,
      "loss": 0.2154,
      "mean_token_accuracy": 0.9456183016300201,
      "num_tokens": 920273.0,
      "step": 122
    },
    {
      "epoch": 1.5605095541401273,
      "grad_norm": 0.4239864945411682,
      "learning_rate": 9.860995952316851e-05,
      "loss": 0.1709,
      "mean_token_accuracy": 0.9573950469493866,
      "num_tokens": 927864.0,
      "step": 123
    },
    {
      "epoch": 1.573248407643312,
      "grad_norm": 0.5215359926223755,
      "learning_rate": 9.722018764467461e-05,
      "loss": 0.1979,
      "mean_token_accuracy": 0.9506862163543701,
      "num_tokens": 935454.0,
      "step": 124
    },
    {
      "epoch": 1.5859872611464967,
      "grad_norm": 0.5375974178314209,
      "learning_rate": 9.583095291095453e-05,
      "loss": 0.2106,
      "mean_token_accuracy": 0.9438005685806274,
      "num_tokens": 943211.0,
      "step": 125
    },
    {
      "epoch": 1.5987261146496814,
      "grad_norm": 0.47071683406829834,
      "learning_rate": 9.444252376465171e-05,
      "loss": 0.188,
      "mean_token_accuracy": 0.9508571922779083,
      "num_tokens": 950616.0,
      "step": 126
    },
    {
      "epoch": 1.611464968152866,
      "grad_norm": 0.4657401442527771,
      "learning_rate": 9.305516849274541e-05,
      "loss": 0.1689,
      "mean_token_accuracy": 0.9533426463603973,
      "num_tokens": 958183.0,
      "step": 127
    },
    {
      "epoch": 1.6242038216560508,
      "grad_norm": 0.3836834132671356,
      "learning_rate": 9.166915517470953e-05,
      "loss": 0.1871,
      "mean_token_accuracy": 0.9511479735374451,
      "num_tokens": 965678.0,
      "step": 128
    },
    {
      "epoch": 1.6369426751592355,
      "grad_norm": 0.46731242537498474,
      "learning_rate": 9.028475163071141e-05,
      "loss": 0.175,
      "mean_token_accuracy": 0.9512114226818085,
      "num_tokens": 973262.0,
      "step": 129
    },
    {
      "epoch": 1.6496815286624202,
      "grad_norm": 0.4541298747062683,
      "learning_rate": 8.890222536986085e-05,
      "loss": 0.216,
      "mean_token_accuracy": 0.9441925585269928,
      "num_tokens": 980887.0,
      "step": 130
    },
    {
      "epoch": 1.662420382165605,
      "grad_norm": 0.3771320581436157,
      "learning_rate": 8.752184353851916e-05,
      "loss": 0.1853,
      "mean_token_accuracy": 0.952815979719162,
      "num_tokens": 988463.0,
      "step": 131
    },
    {
      "epoch": 1.6751592356687897,
      "grad_norm": 0.4476616382598877,
      "learning_rate": 8.614387286867814e-05,
      "loss": 0.1831,
      "mean_token_accuracy": 0.9529510140419006,
      "num_tokens": 996025.0,
      "step": 132
    },
    {
      "epoch": 1.6878980891719744,
      "grad_norm": 0.3788629174232483,
      "learning_rate": 8.47685796264195e-05,
      "loss": 0.194,
      "mean_token_accuracy": 0.953939825296402,
      "num_tokens": 1003816.0,
      "step": 133
    },
    {
      "epoch": 1.700636942675159,
      "grad_norm": 0.4536489248275757,
      "learning_rate": 8.339622956046417e-05,
      "loss": 0.1445,
      "mean_token_accuracy": 0.9615342020988464,
      "num_tokens": 1011351.0,
      "step": 134
    },
    {
      "epoch": 1.7133757961783438,
      "grad_norm": 0.41908058524131775,
      "learning_rate": 8.202708785082121e-05,
      "loss": 0.1623,
      "mean_token_accuracy": 0.9565115869045258,
      "num_tokens": 1018759.0,
      "step": 135
    },
    {
      "epoch": 1.7261146496815285,
      "grad_norm": 0.475737988948822,
      "learning_rate": 8.066141905754723e-05,
      "loss": 0.1725,
      "mean_token_accuracy": 0.9567776024341583,
      "num_tokens": 1026286.0,
      "step": 136
    },
    {
      "epoch": 1.7388535031847132,
      "grad_norm": 0.349752813577652,
      "learning_rate": 7.929948706962508e-05,
      "loss": 0.1255,
      "mean_token_accuracy": 0.9658943712711334,
      "num_tokens": 1033770.0,
      "step": 137
    },
    {
      "epoch": 1.7515923566878981,
      "grad_norm": 0.39415788650512695,
      "learning_rate": 7.794155505397261e-05,
      "loss": 0.1577,
      "mean_token_accuracy": 0.9558201730251312,
      "num_tokens": 1041659.0,
      "step": 138
    },
    {
      "epoch": 1.7643312101910829,
      "grad_norm": 0.41134774684906006,
      "learning_rate": 7.658788540459062e-05,
      "loss": 0.2085,
      "mean_token_accuracy": 0.948514312505722,
      "num_tokens": 1049150.0,
      "step": 139
    },
    {
      "epoch": 1.7770700636942676,
      "grad_norm": 0.5479893088340759,
      "learning_rate": 7.523873969186039e-05,
      "loss": 0.1663,
      "mean_token_accuracy": 0.9547719359397888,
      "num_tokens": 1056784.0,
      "step": 140
    },
    {
      "epoch": 1.7898089171974523,
      "grad_norm": 0.36870864033699036,
      "learning_rate": 7.389437861200024e-05,
      "loss": 0.1485,
      "mean_token_accuracy": 0.9612145721912384,
      "num_tokens": 1064184.0,
      "step": 141
    },
    {
      "epoch": 1.802547770700637,
      "grad_norm": 0.39722350239753723,
      "learning_rate": 7.25550619366911e-05,
      "loss": 0.1559,
      "mean_token_accuracy": 0.9604078829288483,
      "num_tokens": 1072029.0,
      "step": 142
    },
    {
      "epoch": 1.8152866242038217,
      "grad_norm": 0.40050435066223145,
      "learning_rate": 7.122104846288064e-05,
      "loss": 0.1384,
      "mean_token_accuracy": 0.962491363286972,
      "num_tokens": 1079591.0,
      "step": 143
    },
    {
      "epoch": 1.8280254777070064,
      "grad_norm": 0.391933411359787,
      "learning_rate": 6.989259596277582e-05,
      "loss": 0.1792,
      "mean_token_accuracy": 0.9526192545890808,
      "num_tokens": 1087338.0,
      "step": 144
    },
    {
      "epoch": 1.8407643312101911,
      "grad_norm": 0.456823468208313,
      "learning_rate": 6.85699611340333e-05,
      "loss": 0.1416,
      "mean_token_accuracy": 0.9639541804790497,
      "num_tokens": 1094894.0,
      "step": 145
    },
    {
      "epoch": 1.8535031847133758,
      "grad_norm": 0.3839541971683502,
      "learning_rate": 6.725339955015777e-05,
      "loss": 0.1389,
      "mean_token_accuracy": 0.9645723104476929,
      "num_tokens": 1102210.0,
      "step": 146
    },
    {
      "epoch": 1.8662420382165605,
      "grad_norm": 0.33357852697372437,
      "learning_rate": 6.594316561111724e-05,
      "loss": 0.1557,
      "mean_token_accuracy": 0.96068075299263,
      "num_tokens": 1109738.0,
      "step": 147
    },
    {
      "epoch": 1.8789808917197452,
      "grad_norm": 0.4593367874622345,
      "learning_rate": 6.46395124941851e-05,
      "loss": 0.1708,
      "mean_token_accuracy": 0.9570999145507812,
      "num_tokens": 1117265.0,
      "step": 148
    },
    {
      "epoch": 1.89171974522293,
      "grad_norm": 0.36817261576652527,
      "learning_rate": 6.334269210501875e-05,
      "loss": 0.1372,
      "mean_token_accuracy": 0.9625333249568939,
      "num_tokens": 1124990.0,
      "step": 149
    },
    {
      "epoch": 1.9044585987261147,
      "grad_norm": 0.3801482617855072,
      "learning_rate": 6.205295502898348e-05,
      "loss": 0.121,
      "mean_token_accuracy": 0.966850757598877,
      "num_tokens": 1132869.0,
      "step": 150
    },
    {
      "epoch": 1.9171974522292994,
      "grad_norm": 0.34184518456459045,
      "learning_rate": 6.0770550482731924e-05,
      "loss": 0.1642,
      "mean_token_accuracy": 0.9570443630218506,
      "num_tokens": 1140785.0,
      "step": 151
    },
    {
      "epoch": 1.929936305732484,
      "grad_norm": 0.4241127669811249,
      "learning_rate": 5.9495726266047605e-05,
      "loss": 0.1556,
      "mean_token_accuracy": 0.9617596864700317,
      "num_tokens": 1148603.0,
      "step": 152
    },
    {
      "epoch": 1.9426751592356688,
      "grad_norm": 0.3778759837150574,
      "learning_rate": 5.8228728713962543e-05,
      "loss": 0.114,
      "mean_token_accuracy": 0.9691117703914642,
      "num_tokens": 1156149.0,
      "step": 153
    },
    {
      "epoch": 1.9554140127388535,
      "grad_norm": 0.4415481686592102,
      "learning_rate": 5.696980264915777e-05,
      "loss": 0.1381,
      "mean_token_accuracy": 0.9639294743537903,
      "num_tokens": 1163656.0,
      "step": 154
    },
    {
      "epoch": 1.9681528662420382,
      "grad_norm": 0.3667774796485901,
      "learning_rate": 5.571919133465605e-05,
      "loss": 0.1433,
      "mean_token_accuracy": 0.9641910791397095,
      "num_tokens": 1171141.0,
      "step": 155
    },
    {
      "epoch": 1.980891719745223,
      "grad_norm": 0.36915281414985657,
      "learning_rate": 5.447713642681612e-05,
      "loss": 0.1358,
      "mean_token_accuracy": 0.9650273025035858,
      "num_tokens": 1178660.0,
      "step": 156
    },
    {
      "epoch": 1.9936305732484076,
      "grad_norm": 0.33127549290657043,
      "learning_rate": 5.324387792863719e-05,
      "loss": 0.1873,
      "mean_token_accuracy": 0.955519050359726,
      "num_tokens": 1186058.0,
      "step": 157
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5781312584877014,
      "learning_rate": 5.201965414338308e-05,
      "loss": 0.0995,
      "mean_token_accuracy": 0.9694300293922424,
      "num_tokens": 1189950.0,
      "step": 158
    },
    {
      "epoch": 2.0127388535031847,
      "grad_norm": 0.22299824655056,
      "learning_rate": 5.080470162853472e-05,
      "loss": 0.0994,
      "mean_token_accuracy": 0.9725558459758759,
      "num_tokens": 1197547.0,
      "step": 159
    },
    {
      "epoch": 2.0254777070063694,
      "grad_norm": 0.20428019762039185,
      "learning_rate": 4.959925515008002e-05,
      "loss": 0.0981,
      "mean_token_accuracy": 0.9740569591522217,
      "num_tokens": 1205124.0,
      "step": 160
    },
    {
      "epoch": 2.038216560509554,
      "grad_norm": 0.31417369842529297,
      "learning_rate": 4.840354763714991e-05,
      "loss": 0.1128,
      "mean_token_accuracy": 0.9698708951473236,
      "num_tokens": 1212386.0,
      "step": 161
    },
    {
      "epoch": 2.050955414012739,
      "grad_norm": 0.2941860258579254,
      "learning_rate": 4.7217810137009274e-05,
      "loss": 0.1198,
      "mean_token_accuracy": 0.9675701260566711,
      "num_tokens": 1219879.0,
      "step": 162
    },
    {
      "epoch": 2.0636942675159236,
      "grad_norm": 0.22774440050125122,
      "learning_rate": 4.604227177041156e-05,
      "loss": 0.0893,
      "mean_token_accuracy": 0.9721206724643707,
      "num_tokens": 1227344.0,
      "step": 163
    },
    {
      "epoch": 2.0764331210191083,
      "grad_norm": 0.20990870893001556,
      "learning_rate": 4.487715968732568e-05,
      "loss": 0.0897,
      "mean_token_accuracy": 0.9736597836017609,
      "num_tokens": 1234928.0,
      "step": 164
    },
    {
      "epoch": 2.089171974522293,
      "grad_norm": 0.23693329095840454,
      "learning_rate": 4.372269902304363e-05,
      "loss": 0.1029,
      "mean_token_accuracy": 0.9717570245265961,
      "num_tokens": 1242719.0,
      "step": 165
    },
    {
      "epoch": 2.1019108280254777,
      "grad_norm": 0.2804213762283325,
      "learning_rate": 4.257911285467754e-05,
      "loss": 0.1026,
      "mean_token_accuracy": 0.9732820093631744,
      "num_tokens": 1250304.0,
      "step": 166
    },
    {
      "epoch": 2.1146496815286624,
      "grad_norm": 0.28864407539367676,
      "learning_rate": 4.144662215805426e-05,
      "loss": 0.1022,
      "mean_token_accuracy": 0.9715860486030579,
      "num_tokens": 1258105.0,
      "step": 167
    },
    {
      "epoch": 2.127388535031847,
      "grad_norm": 0.3053213059902191,
      "learning_rate": 4.0325445765016145e-05,
      "loss": 0.0995,
      "mean_token_accuracy": 0.9712505340576172,
      "num_tokens": 1265853.0,
      "step": 168
    },
    {
      "epoch": 2.140127388535032,
      "grad_norm": 0.3355003893375397,
      "learning_rate": 3.921580032113602e-05,
      "loss": 0.1006,
      "mean_token_accuracy": 0.9709268808364868,
      "num_tokens": 1273220.0,
      "step": 169
    },
    {
      "epoch": 2.1528662420382165,
      "grad_norm": 0.3259344696998596,
      "learning_rate": 3.8117900243854595e-05,
      "loss": 0.1025,
      "mean_token_accuracy": 0.969818502664566,
      "num_tokens": 1280636.0,
      "step": 170
    },
    {
      "epoch": 2.1656050955414012,
      "grad_norm": 0.25363364815711975,
      "learning_rate": 3.7031957681048604e-05,
      "loss": 0.0899,
      "mean_token_accuracy": 0.973079264163971,
      "num_tokens": 1288501.0,
      "step": 171
    },
    {
      "epoch": 2.178343949044586,
      "grad_norm": 0.33290988206863403,
      "learning_rate": 3.595818247003713e-05,
      "loss": 0.1036,
      "mean_token_accuracy": 0.9701544344425201,
      "num_tokens": 1296181.0,
      "step": 172
    },
    {
      "epoch": 2.1910828025477707,
      "grad_norm": 0.23419451713562012,
      "learning_rate": 3.489678209703475e-05,
      "loss": 0.0922,
      "mean_token_accuracy": 0.9737686514854431,
      "num_tokens": 1303725.0,
      "step": 173
    },
    {
      "epoch": 2.2038216560509554,
      "grad_norm": 0.25606822967529297,
      "learning_rate": 3.3847961657058845e-05,
      "loss": 0.093,
      "mean_token_accuracy": 0.9733033776283264,
      "num_tokens": 1311315.0,
      "step": 174
    },
    {
      "epoch": 2.21656050955414,
      "grad_norm": 0.2730248272418976,
      "learning_rate": 3.281192381429894e-05,
      "loss": 0.0945,
      "mean_token_accuracy": 0.9713809788227081,
      "num_tokens": 1318647.0,
      "step": 175
    },
    {
      "epoch": 2.229299363057325,
      "grad_norm": 0.25905486941337585,
      "learning_rate": 3.178886876295578e-05,
      "loss": 0.094,
      "mean_token_accuracy": 0.9720935225486755,
      "num_tokens": 1326203.0,
      "step": 176
    },
    {
      "epoch": 2.2420382165605095,
      "grad_norm": 0.21770048141479492,
      "learning_rate": 3.077899418855772e-05,
      "loss": 0.095,
      "mean_token_accuracy": 0.972693145275116,
      "num_tokens": 1333736.0,
      "step": 177
    },
    {
      "epoch": 2.254777070063694,
      "grad_norm": 0.21615815162658691,
      "learning_rate": 2.9782495229761808e-05,
      "loss": 0.0959,
      "mean_token_accuracy": 0.9725539982318878,
      "num_tokens": 1341160.0,
      "step": 178
    },
    {
      "epoch": 2.267515923566879,
      "grad_norm": 0.29113566875457764,
      "learning_rate": 2.879956444064703e-05,
      "loss": 0.1089,
      "mean_token_accuracy": 0.9702370762825012,
      "num_tokens": 1348750.0,
      "step": 179
    },
    {
      "epoch": 2.2802547770700636,
      "grad_norm": 0.298750102519989,
      "learning_rate": 2.783039175350699e-05,
      "loss": 0.1037,
      "mean_token_accuracy": 0.9692679643630981,
      "num_tokens": 1356569.0,
      "step": 180
    },
    {
      "epoch": 2.2929936305732483,
      "grad_norm": 0.2231506109237671,
      "learning_rate": 2.6875164442149147e-05,
      "loss": 0.0955,
      "mean_token_accuracy": 0.9708955585956573,
      "num_tokens": 1364056.0,
      "step": 181
    },
    {
      "epoch": 2.305732484076433,
      "grad_norm": 0.2941678762435913,
      "learning_rate": 2.5934067085707834e-05,
      "loss": 0.0996,
      "mean_token_accuracy": 0.9709564447402954,
      "num_tokens": 1371659.0,
      "step": 182
    },
    {
      "epoch": 2.3184713375796178,
      "grad_norm": 0.1835247129201889,
      "learning_rate": 2.500728153297788e-05,
      "loss": 0.0942,
      "mean_token_accuracy": 0.9725345969200134,
      "num_tokens": 1379147.0,
      "step": 183
    },
    {
      "epoch": 2.3312101910828025,
      "grad_norm": 0.28575223684310913,
      "learning_rate": 2.409498686727587e-05,
      "loss": 0.091,
      "mean_token_accuracy": 0.9733802676200867,
      "num_tokens": 1386765.0,
      "step": 184
    },
    {
      "epoch": 2.343949044585987,
      "grad_norm": 0.2463427037000656,
      "learning_rate": 2.3197359371835802e-05,
      "loss": 0.086,
      "mean_token_accuracy": 0.9742932319641113,
      "num_tokens": 1394618.0,
      "step": 185
    },
    {
      "epoch": 2.356687898089172,
      "grad_norm": 0.1932964026927948,
      "learning_rate": 2.2314572495745746e-05,
      "loss": 0.093,
      "mean_token_accuracy": 0.9733997285366058,
      "num_tokens": 1402323.0,
      "step": 186
    },
    {
      "epoch": 2.3694267515923566,
      "grad_norm": 0.31327489018440247,
      "learning_rate": 2.1446796820432167e-05,
      "loss": 0.099,
      "mean_token_accuracy": 0.9692893028259277,
      "num_tokens": 1409811.0,
      "step": 187
    },
    {
      "epoch": 2.3821656050955413,
      "grad_norm": 0.2381305992603302,
      "learning_rate": 2.0594200026698363e-05,
      "loss": 0.0944,
      "mean_token_accuracy": 0.9724844992160797,
      "num_tokens": 1417360.0,
      "step": 188
    },
    {
      "epoch": 2.394904458598726,
      "grad_norm": 0.2520155906677246,
      "learning_rate": 1.9756946862323535e-05,
      "loss": 0.096,
      "mean_token_accuracy": 0.9740960896015167,
      "num_tokens": 1425115.0,
      "step": 189
    },
    {
      "epoch": 2.4076433121019107,
      "grad_norm": 0.183822900056839,
      "learning_rate": 1.8935199110228275e-05,
      "loss": 0.0911,
      "mean_token_accuracy": 0.9735857248306274,
      "num_tokens": 1432795.0,
      "step": 190
    },
    {
      "epoch": 2.4203821656050954,
      "grad_norm": 0.20515495538711548,
      "learning_rate": 1.8129115557213262e-05,
      "loss": 0.0936,
      "mean_token_accuracy": 0.9743000864982605,
      "num_tokens": 1440406.0,
      "step": 191
    },
    {
      "epoch": 2.43312101910828,
      "grad_norm": 0.34444665908813477,
      "learning_rate": 1.7338851963276825e-05,
      "loss": 0.0932,
      "mean_token_accuracy": 0.9733559787273407,
      "num_tokens": 1447979.0,
      "step": 192
    },
    {
      "epoch": 2.445859872611465,
      "grad_norm": 0.19695109128952026,
      "learning_rate": 1.656456103151728e-05,
      "loss": 0.0941,
      "mean_token_accuracy": 0.9739238917827606,
      "num_tokens": 1455491.0,
      "step": 193
    },
    {
      "epoch": 2.4585987261146496,
      "grad_norm": 0.21720701456069946,
      "learning_rate": 1.580639237862608e-05,
      "loss": 0.0928,
      "mean_token_accuracy": 0.9730100631713867,
      "num_tokens": 1463007.0,
      "step": 194
    },
    {
      "epoch": 2.4713375796178343,
      "grad_norm": 0.21964755654335022,
      "learning_rate": 1.5064492505977234e-05,
      "loss": 0.0892,
      "mean_token_accuracy": 0.9735862016677856,
      "num_tokens": 1470901.0,
      "step": 195
    },
    {
      "epoch": 2.484076433121019,
      "grad_norm": 0.1945929080247879,
      "learning_rate": 1.433900477131882e-05,
      "loss": 0.0889,
      "mean_token_accuracy": 0.9738577008247375,
      "num_tokens": 1478538.0,
      "step": 196
    },
    {
      "epoch": 2.4968152866242037,
      "grad_norm": 0.2573101818561554,
      "learning_rate": 1.363006936107183e-05,
      "loss": 0.1038,
      "mean_token_accuracy": 0.9701802432537079,
      "num_tokens": 1486088.0,
      "step": 197
    },
    {
      "epoch": 2.5095541401273884,
      "grad_norm": 0.1948987990617752,
      "learning_rate": 1.29378232632419e-05,
      "loss": 0.0892,
      "mean_token_accuracy": 0.9724543392658234,
      "num_tokens": 1493703.0,
      "step": 198
    },
    {
      "epoch": 2.522292993630573,
      "grad_norm": 0.2603251338005066,
      "learning_rate": 1.2262400240949023e-05,
      "loss": 0.1021,
      "mean_token_accuracy": 0.9714818596839905,
      "num_tokens": 1501201.0,
      "step": 199
    },
    {
      "epoch": 2.535031847133758,
      "grad_norm": 0.23355098068714142,
      "learning_rate": 1.1603930806580444e-05,
      "loss": 0.0918,
      "mean_token_accuracy": 0.9732993841171265,
      "num_tokens": 1508923.0,
      "step": 200
    },
    {
      "epoch": 2.5477707006369426,
      "grad_norm": 0.26143762469291687,
      "learning_rate": 1.0962542196571634e-05,
      "loss": 0.0978,
      "mean_token_accuracy": 0.9724924564361572,
      "num_tokens": 1516403.0,
      "step": 201
    },
    {
      "epoch": 2.5605095541401273,
      "grad_norm": 0.19722925126552582,
      "learning_rate": 1.0338358346820353e-05,
      "loss": 0.0881,
      "mean_token_accuracy": 0.9745533466339111,
      "num_tokens": 1524170.0,
      "step": 202
    },
    {
      "epoch": 2.573248407643312,
      "grad_norm": 0.24879567325115204,
      "learning_rate": 9.731499868738447e-06,
      "loss": 0.1024,
      "mean_token_accuracy": 0.9704383015632629,
      "num_tokens": 1531711.0,
      "step": 203
    },
    {
      "epoch": 2.5859872611464967,
      "grad_norm": 0.18551646173000336,
      "learning_rate": 9.142084025945984e-06,
      "loss": 0.0887,
      "mean_token_accuracy": 0.9731804728507996,
      "num_tokens": 1539309.0,
      "step": 204
    },
    {
      "epoch": 2.5987261146496814,
      "grad_norm": 0.19061441719532013,
      "learning_rate": 8.570224711612385e-06,
      "loss": 0.0906,
      "mean_token_accuracy": 0.9726877510547638,
      "num_tokens": 1546875.0,
      "step": 205
    },
    {
      "epoch": 2.611464968152866,
      "grad_norm": 0.22261416912078857,
      "learning_rate": 8.016032426448817e-06,
      "loss": 0.0918,
      "mean_token_accuracy": 0.9743273258209229,
      "num_tokens": 1554385.0,
      "step": 206
    },
    {
      "epoch": 2.624203821656051,
      "grad_norm": 0.23906444013118744,
      "learning_rate": 7.479614257355971e-06,
      "loss": 0.0954,
      "mean_token_accuracy": 0.9728935956954956,
      "num_tokens": 1562049.0,
      "step": 207
    },
    {
      "epoch": 2.6369426751592355,
      "grad_norm": 0.2492460012435913,
      "learning_rate": 6.961073856731648e-06,
      "loss": 0.096,
      "mean_token_accuracy": 0.9732304513454437,
      "num_tokens": 1569434.0,
      "step": 208
    },
    {
      "epoch": 2.6496815286624202,
      "grad_norm": 0.278803288936615,
      "learning_rate": 6.460511422441984e-06,
      "loss": 0.0969,
      "mean_token_accuracy": 0.9719431698322296,
      "num_tokens": 1576927.0,
      "step": 209
    },
    {
      "epoch": 2.662420382165605,
      "grad_norm": 0.24254080653190613,
      "learning_rate": 5.978023678460099e-06,
      "loss": 0.1048,
      "mean_token_accuracy": 0.9709457159042358,
      "num_tokens": 1584531.0,
      "step": 210
    },
    {
      "epoch": 2.6751592356687897,
      "grad_norm": 0.21079742908477783,
      "learning_rate": 5.5137038561761115e-06,
      "loss": 0.0892,
      "mean_token_accuracy": 0.9735644161701202,
      "num_tokens": 1591932.0,
      "step": 211
    },
    {
      "epoch": 2.6878980891719744,
      "grad_norm": 0.25342243909835815,
      "learning_rate": 5.067641676381918e-06,
      "loss": 0.0972,
      "mean_token_accuracy": 0.972582995891571,
      "num_tokens": 1599678.0,
      "step": 212
    },
    {
      "epoch": 2.700636942675159,
      "grad_norm": 0.1848929524421692,
      "learning_rate": 4.639923331934471e-06,
      "loss": 0.0904,
      "mean_token_accuracy": 0.97327721118927,
      "num_tokens": 1607234.0,
      "step": 213
    },
    {
      "epoch": 2.713375796178344,
      "grad_norm": 0.21844594180583954,
      "learning_rate": 4.230631471100655e-06,
      "loss": 0.098,
      "mean_token_accuracy": 0.9728173017501831,
      "num_tokens": 1614910.0,
      "step": 214
    },
    {
      "epoch": 2.7261146496815285,
      "grad_norm": 0.29655906558036804,
      "learning_rate": 3.839845181587098e-06,
      "loss": 0.1011,
      "mean_token_accuracy": 0.9714401960372925,
      "num_tokens": 1622358.0,
      "step": 215
    },
    {
      "epoch": 2.738853503184713,
      "grad_norm": 0.19944727420806885,
      "learning_rate": 3.467639975257997e-06,
      "loss": 0.0913,
      "mean_token_accuracy": 0.9719055891036987,
      "num_tokens": 1629753.0,
      "step": 216
    },
    {
      "epoch": 2.7515923566878984,
      "grad_norm": 0.21076829731464386,
      "learning_rate": 3.1140877735439387e-06,
      "loss": 0.0894,
      "mean_token_accuracy": 0.97262904047966,
      "num_tokens": 1637347.0,
      "step": 217
    },
    {
      "epoch": 2.7643312101910826,
      "grad_norm": 0.2960074245929718,
      "learning_rate": 2.7792568935444796e-06,
      "loss": 0.0913,
      "mean_token_accuracy": 0.9721934199333191,
      "num_tokens": 1644999.0,
      "step": 218
    },
    {
      "epoch": 2.777070063694268,
      "grad_norm": 0.185651496052742,
      "learning_rate": 2.4632120348272003e-06,
      "loss": 0.0888,
      "mean_token_accuracy": 0.9723814725875854,
      "num_tokens": 1652451.0,
      "step": 219
    },
    {
      "epoch": 2.789808917197452,
      "grad_norm": 0.22099068760871887,
      "learning_rate": 2.166014266925731e-06,
      "loss": 0.0892,
      "mean_token_accuracy": 0.9731795489788055,
      "num_tokens": 1660101.0,
      "step": 220
    },
    {
      "epoch": 2.802547770700637,
      "grad_norm": 0.2245907336473465,
      "learning_rate": 1.88772101753929e-06,
      "loss": 0.0939,
      "mean_token_accuracy": 0.9741968810558319,
      "num_tokens": 1667978.0,
      "step": 221
    },
    {
      "epoch": 2.8152866242038215,
      "grad_norm": 0.2256772518157959,
      "learning_rate": 1.6283860614358936e-06,
      "loss": 0.0914,
      "mean_token_accuracy": 0.9728004634380341,
      "num_tokens": 1675469.0,
      "step": 222
    },
    {
      "epoch": 2.8280254777070066,
      "grad_norm": 0.19644984602928162,
      "learning_rate": 1.3880595100613792e-06,
      "loss": 0.0927,
      "mean_token_accuracy": 0.9726226031780243,
      "num_tokens": 1682842.0,
      "step": 223
    },
    {
      "epoch": 2.840764331210191,
      "grad_norm": 0.22729112207889557,
      "learning_rate": 1.1667878018564171e-06,
      "loss": 0.0903,
      "mean_token_accuracy": 0.9728392660617828,
      "num_tokens": 1690491.0,
      "step": 224
    },
    {
      "epoch": 2.853503184713376,
      "grad_norm": 0.21743939816951752,
      "learning_rate": 9.64613693283123e-07,
      "loss": 0.0956,
      "mean_token_accuracy": 0.9711977541446686,
      "num_tokens": 1697950.0,
      "step": 225
    },
    {
      "epoch": 2.8662420382165603,
      "grad_norm": 0.23131786286830902,
      "learning_rate": 7.815762505632096e-07,
      "loss": 0.0908,
      "mean_token_accuracy": 0.972371518611908,
      "num_tokens": 1705398.0,
      "step": 226
    },
    {
      "epoch": 2.8789808917197455,
      "grad_norm": 0.19186514616012573,
      "learning_rate": 6.177108421292266e-07,
      "loss": 0.0896,
      "mean_token_accuracy": 0.9753928780555725,
      "num_tokens": 1712737.0,
      "step": 227
    },
    {
      "epoch": 2.8917197452229297,
      "grad_norm": 0.15185010433197021,
      "learning_rate": 4.7304913179025965e-07,
      "loss": 0.0849,
      "mean_token_accuracy": 0.9748527407646179,
      "num_tokens": 1720472.0,
      "step": 228
    },
    {
      "epoch": 2.904458598726115,
      "grad_norm": 0.1807604730129242,
      "learning_rate": 3.4761907261356976e-07,
      "loss": 0.0887,
      "mean_token_accuracy": 0.9729354083538055,
      "num_tokens": 1727956.0,
      "step": 229
    },
    {
      "epoch": 2.917197452229299,
      "grad_norm": 0.23020492494106293,
      "learning_rate": 2.414449015231357e-07,
      "loss": 0.0956,
      "mean_token_accuracy": 0.9716013371944427,
      "num_tokens": 1735520.0,
      "step": 230
    },
    {
      "epoch": 2.9299363057324843,
      "grad_norm": 0.3724834620952606,
      "learning_rate": 1.545471346164007e-07,
      "loss": 0.1087,
      "mean_token_accuracy": 0.9690882861614227,
      "num_tokens": 1743213.0,
      "step": 231
    },
    {
      "epoch": 2.9426751592356686,
      "grad_norm": 0.3106917142868042,
      "learning_rate": 8.694256319987659e-08,
      "loss": 0.0969,
      "mean_token_accuracy": 0.9719716310501099,
      "num_tokens": 1750982.0,
      "step": 232
    },
    {
      "epoch": 2.9554140127388537,
      "grad_norm": 0.19371703267097473,
      "learning_rate": 3.8644250544594975e-08,
      "loss": 0.0889,
      "mean_token_accuracy": 0.972972184419632,
      "num_tokens": 1758448.0,
      "step": 233
    },
    {
      "epoch": 2.968152866242038,
      "grad_norm": 0.29524293541908264,
      "learning_rate": 9.661529361892907e-09,
      "loss": 0.0987,
      "mean_token_accuracy": 0.9689370095729828,
      "num_tokens": 1766023.0,
      "step": 234
    },
    {
      "epoch": 2.968152866242038,
      "step": 234,
      "total_flos": 1.0230160156105114e+17,
      "train_loss": 0.3221458565985036,
      "train_runtime": 403.0093,
      "train_samples_per_second": 37.22,
      "train_steps_per_second": 0.581
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 234,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0230160156105114e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}