{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 1024,
  "global_step": 8100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06320987654320988,
      "grad_norm": 0.15384447574615479,
      "learning_rate": 0.000248046875,
      "loss": 0.17836107313632965,
      "step": 128
    },
    {
      "epoch": 0.12641975308641976,
      "grad_norm": 0.19778844714164734,
      "learning_rate": 0.000498046875,
      "loss": 0.0778987780213356,
      "step": 256
    },
    {
      "epoch": 0.18962962962962962,
      "grad_norm": 0.1558559685945511,
      "learning_rate": 0.000748046875,
      "loss": 0.07612956315279007,
      "step": 384
    },
    {
      "epoch": 0.2528395061728395,
      "grad_norm": 0.20315754413604736,
      "learning_rate": 0.000998046875,
      "loss": 0.07527799159288406,
      "step": 512
    },
    {
      "epoch": 0.3160493827160494,
      "grad_norm": 0.23633696138858795,
      "learning_rate": 0.0009993089770195807,
      "loss": 0.07513153553009033,
      "step": 640
    },
    {
      "epoch": 0.37925925925925924,
      "grad_norm": 0.18408338725566864,
      "learning_rate": 0.0009972160460972733,
      "loss": 0.07528101652860641,
      "step": 768
    },
    {
      "epoch": 0.44246913580246916,
      "grad_norm": 0.07793154567480087,
      "learning_rate": 0.0009937270408736224,
      "loss": 0.07278512418270111,
      "step": 896
    },
    {
      "epoch": 0.505679012345679,
      "grad_norm": 0.05981123447418213,
      "learning_rate": 0.0009888517577149526,
      "loss": 0.07673695683479309,
      "step": 1024
    },
    {
      "epoch": 0.505679012345679,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.025346231987563576,
      "eval_loss": 0.07089577915101516,
      "eval_mse_loss": 0.07089577915101516,
      "step": 1024
    },
    {
      "epoch": 0.505679012345679,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.025346231987563576,
      "eval_loss": 0.07089577915101516,
      "eval_mse_loss": 0.07089577915101516,
      "eval_runtime": 17.586,
      "eval_samples_per_second": 148.868,
      "eval_steps_per_second": 2.331,
      "step": 1024
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 0.1194850280880928,
      "learning_rate": 0.0009826038853539248,
      "loss": 0.07441737502813339,
      "step": 1152
    },
    {
      "epoch": 0.6320987654320988,
      "grad_norm": 0.2225128412246704,
      "learning_rate": 0.0009750009664545572,
      "loss": 0.07503814250230789,
      "step": 1280
    },
    {
      "epoch": 0.6953086419753086,
      "grad_norm": 0.08698221296072006,
      "learning_rate": 0.0009660643483562486,
      "loss": 0.07423833012580872,
      "step": 1408
    },
    {
      "epoch": 0.7585185185185185,
      "grad_norm": 0.07486765086650848,
      "learning_rate": 0.0009558191231351013,
      "loss": 0.07421601563692093,
      "step": 1536
    },
    {
      "epoch": 0.8217283950617283,
      "grad_norm": 0.048349592834711075,
      "learning_rate": 0.0009442940571508399,
      "loss": 0.07361512631177902,
      "step": 1664
    },
    {
      "epoch": 0.8849382716049383,
      "grad_norm": 0.05034460127353668,
      "learning_rate": 0.0009315215102771411,
      "loss": 0.07331028580665588,
      "step": 1792
    },
    {
      "epoch": 0.9481481481481482,
      "grad_norm": 0.05456356704235077,
      "learning_rate": 0.0009175373450421618,
      "loss": 0.07391860336065292,
      "step": 1920
    },
    {
      "epoch": 1.011358024691358,
      "grad_norm": 0.06820514798164368,
      "learning_rate": 0.0009023808259343743,
      "loss": 0.07479690760374069,
      "step": 2048
    },
    {
      "epoch": 1.011358024691358,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.026905295779792275,
      "eval_loss": 0.07202080500925459,
      "eval_mse_loss": 0.07202080500925459,
      "step": 2048
    },
    {
      "epoch": 1.011358024691358,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.026905295779792275,
      "eval_loss": 0.07202080500925459,
      "eval_mse_loss": 0.07202080500925459,
      "eval_runtime": 17.8305,
      "eval_samples_per_second": 146.827,
      "eval_steps_per_second": 2.299,
      "step": 2048
    },
    {
      "epoch": 1.074567901234568,
      "grad_norm": 0.09580802917480469,
      "learning_rate": 0.0008860945091564399,
      "loss": 0.07324858009815216,
      "step": 2176
    },
    {
      "epoch": 1.1377777777777778,
      "grad_norm": 0.07308576256036758,
      "learning_rate": 0.0008687241231366662,
      "loss": 0.07369088381528854,
      "step": 2304
    },
    {
      "epoch": 1.2009876543209876,
      "grad_norm": 0.262474924325943,
      "learning_rate": 0.0008503184401335448,
      "loss": 0.07218585908412933,
      "step": 2432
    },
    {
      "epoch": 1.2641975308641975,
      "grad_norm": 0.06500956416130066,
      "learning_rate": 0.0008309291392938795,
      "loss": 0.07315586507320404,
      "step": 2560
    },
    {
      "epoch": 1.3274074074074074,
      "grad_norm": 0.06056557968258858,
      "learning_rate": 0.0008106106615490032,
      "loss": 0.0738791897892952,
      "step": 2688
    },
    {
      "epoch": 1.3906172839506172,
      "grad_norm": 0.0900755524635315,
      "learning_rate": 0.0007894200567565075,
      "loss": 0.07376813143491745,
      "step": 2816
    },
    {
      "epoch": 1.453827160493827,
      "grad_norm": 0.05714402347803116,
      "learning_rate": 0.0007674168235166747,
      "loss": 0.07344707101583481,
      "step": 2944
    },
    {
      "epoch": 1.5170370370370372,
      "grad_norm": 0.10205268859863281,
      "learning_rate": 0.000744662742113375,
      "loss": 0.07367056608200073,
      "step": 3072
    },
    {
      "epoch": 1.5170370370370372,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.02579855146567996,
      "eval_loss": 0.0767637947347106,
      "eval_mse_loss": 0.0767637947347106,
      "step": 3072
    },
    {
      "epoch": 1.5170370370370372,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.02579855146567996,
      "eval_loss": 0.0767637947347106,
      "eval_mse_loss": 0.0767637947347106,
      "eval_runtime": 17.4674,
      "eval_samples_per_second": 149.879,
      "eval_steps_per_second": 2.347,
      "step": 3072
    },
    {
      "epoch": 1.5802469135802468,
      "grad_norm": 0.5726671814918518,
      "learning_rate": 0.0007212217010484917,
      "loss": 0.07431654632091522,
      "step": 3200
    },
    {
      "epoch": 1.643456790123457,
      "grad_norm": 0.15511786937713623,
      "learning_rate": 0.0006971595176569332,
      "loss": 0.07296353578567505,
      "step": 3328
    },
    {
      "epoch": 1.7066666666666666,
      "grad_norm": 0.05120816081762314,
      "learning_rate": 0.0006725437533058971,
      "loss": 0.07412794977426529,
      "step": 3456
    },
    {
      "epoch": 1.7698765432098766,
      "grad_norm": 0.05225963890552521,
      "learning_rate": 0.0006474435236972767,
      "loss": 0.07479549944400787,
      "step": 3584
    },
    {
      "epoch": 1.8330864197530863,
      "grad_norm": 0.08605295419692993,
      "learning_rate": 0.0006219293048058301,
      "loss": 0.07417000830173492,
      "step": 3712
    },
    {
      "epoch": 1.8962962962962964,
      "grad_norm": 0.07357269525527954,
      "learning_rate": 0.0005960727349980042,
      "loss": 0.07279553264379501,
      "step": 3840
    },
    {
      "epoch": 1.959506172839506,
      "grad_norm": 0.16880673170089722,
      "learning_rate": 0.0005699464138870106,
      "loss": 0.07310041785240173,
      "step": 3968
    },
    {
      "epoch": 2.022716049382716,
      "grad_norm": 0.05540848150849342,
      "learning_rate": 0.0005436236984889356,
      "loss": 0.07398485392332077,
      "step": 4096
    },
    {
      "epoch": 2.022716049382716,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.025484831140535635,
      "eval_loss": 0.07245415630863934,
      "eval_mse_loss": 0.07245415630863934,
      "step": 4096
    },
    {
      "epoch": 2.022716049382716,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.025484831140535635,
      "eval_loss": 0.07245415630863934,
      "eval_mse_loss": 0.07245415630863934,
      "eval_runtime": 17.6066,
      "eval_samples_per_second": 148.694,
      "eval_steps_per_second": 2.329,
      "step": 4096
    },
    {
      "epoch": 2.0859259259259257,
      "grad_norm": 0.04602835327386856,
      "learning_rate": 0.0005171784972522235,
      "loss": 0.07218825817108154,
      "step": 4224
    },
    {
      "epoch": 2.149135802469136,
      "grad_norm": 0.04208948090672493,
      "learning_rate": 0.0004906850625388614,
      "loss": 0.07182744145393372,
      "step": 4352
    },
    {
      "epoch": 2.212345679012346,
      "grad_norm": 0.08561620116233826,
      "learning_rate": 0.0004642177821399269,
      "loss": 0.0734778568148613,
      "step": 4480
    },
    {
      "epoch": 2.2755555555555556,
      "grad_norm": 0.03911186754703522,
      "learning_rate": 0.00043785097041088293,
      "loss": 0.07363104820251465,
      "step": 4608
    },
    {
      "epoch": 2.338765432098765,
      "grad_norm": 0.04363245889544487,
      "learning_rate": 0.00041165865961306135,
      "loss": 0.07383761554956436,
      "step": 4736
    },
    {
      "epoch": 2.4019753086419753,
      "grad_norm": 0.05380804464221001,
      "learning_rate": 0.000385714392047208,
      "loss": 0.07362533360719681,
      "step": 4864
    },
    {
      "epoch": 2.4651851851851854,
      "grad_norm": 0.05036137253046036,
      "learning_rate": 0.00036009101356272814,
      "loss": 0.07132267206907272,
      "step": 4992
    },
    {
      "epoch": 2.528395061728395,
      "grad_norm": 0.04343096911907196,
      "learning_rate": 0.00033486046902241664,
      "loss": 0.07365947216749191,
      "step": 5120
    },
    {
      "epoch": 2.528395061728395,
      "eval_bleu": 0.00011770260770184794,
      "eval_cos_loss": 0.024220028346995028,
      "eval_loss": 0.0705922639406309,
      "eval_mse_loss": 0.0705922639406309,
      "step": 5120
    },
    {
      "epoch": 2.528395061728395,
      "eval_bleu": 0.00011770260770184794,
      "eval_cos_loss": 0.024220028346995028,
      "eval_loss": 0.0705922639406309,
      "eval_mse_loss": 0.0705922639406309,
      "eval_runtime": 18.0957,
      "eval_samples_per_second": 144.675,
      "eval_steps_per_second": 2.266,
      "step": 5120
    },
    {
      "epoch": 2.591604938271605,
      "grad_norm": 0.04439703747630119,
      "learning_rate": 0.00031009360029696107,
      "loss": 0.07380591332912445,
      "step": 5248
    },
    {
      "epoch": 2.6548148148148147,
      "grad_norm": 0.10452727228403091,
      "learning_rate": 0.00028585994735640823,
      "loss": 0.07314619421958923,
      "step": 5376
    },
    {
      "epoch": 2.718024691358025,
      "grad_norm": 0.029365811496973038,
      "learning_rate": 0.0002622275530170825,
      "loss": 0.07315413653850555,
      "step": 5504
    },
    {
      "epoch": 2.7812345679012345,
      "grad_norm": 0.026572152972221375,
      "learning_rate": 0.0002392627718921852,
      "loss": 0.07188510149717331,
      "step": 5632
    },
    {
      "epoch": 2.8444444444444446,
      "grad_norm": 0.048810794949531555,
      "learning_rate": 0.00021703008408250187,
      "loss": 0.07245349138975143,
      "step": 5760
    },
    {
      "epoch": 2.907654320987654,
      "grad_norm": 0.03271602466702461,
      "learning_rate": 0.00019559191413033017,
      "loss": 0.072321817278862,
      "step": 5888
    },
    {
      "epoch": 2.9708641975308643,
      "grad_norm": 0.06507620960474014,
      "learning_rate": 0.00017500845574497032,
      "loss": 0.07270237058401108,
      "step": 6016
    },
    {
      "epoch": 3.034074074074074,
      "grad_norm": 0.04906519874930382,
      "learning_rate": 0.00015533750279190617,
      "loss": 0.07204872369766235,
      "step": 6144
    },
    {
      "epoch": 3.034074074074074,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.02610533970703439,
      "eval_loss": 0.07495295165515528,
      "eval_mse_loss": 0.07495295165515528,
      "step": 6144
    },
    {
      "epoch": 3.034074074074074,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.02610533970703439,
      "eval_loss": 0.07495295165515528,
      "eval_mse_loss": 0.07495295165515528,
      "eval_runtime": 18.2816,
      "eval_samples_per_second": 143.204,
      "eval_steps_per_second": 2.243,
      "step": 6144
    },
    {
      "epoch": 3.097283950617284,
      "grad_norm": 0.2643573582172394,
      "learning_rate": 0.000136634287020226,
      "loss": 0.07238763570785522,
      "step": 6272
    },
    {
      "epoch": 3.1604938271604937,
      "grad_norm": 0.02101626992225647,
      "learning_rate": 0.00011895132298390743,
      "loss": 0.07245352864265442,
      "step": 6400
    },
    {
      "epoch": 3.2237037037037037,
      "grad_norm": 0.04818721115589142,
      "learning_rate": 0.00010233826059239426,
      "loss": 0.0718931034207344,
      "step": 6528
    },
    {
      "epoch": 3.286913580246914,
      "grad_norm": 0.031206775456666946,
      "learning_rate": 8.68417457044704e-05,
      "loss": 0.07213378697633743,
      "step": 6656
    },
    {
      "epoch": 3.3501234567901235,
      "grad_norm": 0.04356447234749794,
      "learning_rate": 7.250528915685422e-05,
      "loss": 0.07157152891159058,
      "step": 6784
    },
    {
      "epoch": 3.413333333333333,
      "grad_norm": 0.025733737275004387,
      "learning_rate": 5.93691445952505e-05,
      "loss": 0.07190953195095062,
      "step": 6912
    },
    {
      "epoch": 3.476543209876543,
      "grad_norm": 0.032584454864263535,
      "learning_rate": 4.7470195450886376e-05,
      "loss": 0.07134460657835007,
      "step": 7040
    },
    {
      "epoch": 3.5397530864197533,
      "grad_norm": 0.02684413455426693,
      "learning_rate": 3.684185137987378e-05,
      "loss": 0.07240204513072968,
      "step": 7168
    },
    {
      "epoch": 3.5397530864197533,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.02353674954757458,
      "eval_loss": 0.06911279033960366,
      "eval_mse_loss": 0.06911279033960366,
      "step": 7168
    },
    {
      "epoch": 3.5397530864197533,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.02353674954757458,
      "eval_loss": 0.06911279033960366,
      "eval_mse_loss": 0.06911279033960366,
      "eval_runtime": 19.2347,
      "eval_samples_per_second": 136.108,
      "eval_steps_per_second": 2.132,
      "step": 7168
    },
    {
      "epoch": 3.602962962962963,
      "grad_norm": 0.023628532886505127,
      "learning_rate": 2.751395445617594e-05,
      "loss": 0.07160157710313797,
      "step": 7296
    },
    {
      "epoch": 3.6661728395061726,
      "grad_norm": 0.05646843463182449,
      "learning_rate": 1.9512695381567302e-05,
      "loss": 0.07188498228788376,
      "step": 7424
    },
    {
      "epoch": 3.7293827160493827,
      "grad_norm": 0.04089902713894844,
      "learning_rate": 1.2860539947850303e-05,
      "loss": 0.07162626832723618,
      "step": 7552
    },
    {
      "epoch": 3.7925925925925927,
      "grad_norm": 0.02128681354224682,
      "learning_rate": 7.5761659578078565e-06,
      "loss": 0.07193399965763092,
      "step": 7680
    },
    {
      "epoch": 3.8558024691358024,
      "grad_norm": 0.06368973851203918,
      "learning_rate": 3.674410782003812e-06,
      "loss": 0.07139615714550018,
      "step": 7808
    },
    {
      "epoch": 3.9190123456790125,
      "grad_norm": 0.036761000752449036,
      "learning_rate": 1.1662296986795217e-06,
      "loss": 0.07149382680654526,
      "step": 7936
    },
    {
      "epoch": 3.982222222222222,
      "grad_norm": 0.02737569995224476,
      "learning_rate": 5.866513372004834e-08,
      "loss": 0.07145236432552338,
      "step": 8064
    }
  ],
  "logging_steps": 128,
  "max_steps": 8100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 1024,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}