{
  "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": 0.6517145037651062,
      "learning_rate": 0.000248046875,
      "loss": 0.17245662212371826,
      "step": 128
    },
    {
      "epoch": 0.12641975308641976,
      "grad_norm": 0.2118282914161682,
      "learning_rate": 0.000498046875,
      "loss": 0.15388090908527374,
      "step": 256
    },
    {
      "epoch": 0.18962962962962962,
      "grad_norm": 0.23596829175949097,
      "learning_rate": 0.000748046875,
      "loss": 0.1529599130153656,
      "step": 384
    },
    {
      "epoch": 0.2528395061728395,
      "grad_norm": 0.16465483605861664,
      "learning_rate": 0.000998046875,
      "loss": 0.15048745274543762,
      "step": 512
    },
    {
      "epoch": 0.3160493827160494,
      "grad_norm": 0.09756451845169067,
      "learning_rate": 0.0009993089770195807,
      "loss": 0.1506166309118271,
      "step": 640
    },
    {
      "epoch": 0.37925925925925924,
      "grad_norm": 0.2256879061460495,
      "learning_rate": 0.0009972160460972733,
      "loss": 0.15044106543064117,
      "step": 768
    },
    {
      "epoch": 0.44246913580246916,
      "grad_norm": 0.11761642247438431,
      "learning_rate": 0.0009937270408736224,
      "loss": 0.1501142978668213,
      "step": 896
    },
    {
      "epoch": 0.505679012345679,
      "grad_norm": 0.0816347524523735,
      "learning_rate": 0.0009888517577149526,
      "loss": 0.15066105127334595,
      "step": 1024
    },
    {
      "epoch": 0.505679012345679,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03857117618729428,
      "eval_loss": 0.148995438000051,
      "eval_mse_loss": 0.148995438000051,
      "step": 1024
    },
    {
      "epoch": 0.505679012345679,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03857117618729428,
      "eval_loss": 0.148995438000051,
      "eval_mse_loss": 0.148995438000051,
      "eval_runtime": 17.631,
      "eval_samples_per_second": 148.488,
      "eval_steps_per_second": 2.325,
      "step": 1024
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 0.08801733702421188,
      "learning_rate": 0.0009826038853539248,
      "loss": 0.1511014848947525,
      "step": 1152
    },
    {
      "epoch": 0.6320987654320988,
      "grad_norm": 0.07067783921957016,
      "learning_rate": 0.0009750009664545572,
      "loss": 0.1478763073682785,
      "step": 1280
    },
    {
      "epoch": 0.6953086419753086,
      "grad_norm": 0.08224646747112274,
      "learning_rate": 0.0009660643483562486,
      "loss": 0.15107029676437378,
      "step": 1408
    },
    {
      "epoch": 0.7585185185185185,
      "grad_norm": 0.03443925082683563,
      "learning_rate": 0.0009558191231351013,
      "loss": 0.14998815953731537,
      "step": 1536
    },
    {
      "epoch": 0.8217283950617283,
      "grad_norm": 0.06813549995422363,
      "learning_rate": 0.0009442940571508399,
      "loss": 0.14969073235988617,
      "step": 1664
    },
    {
      "epoch": 0.8849382716049383,
      "grad_norm": 0.059174299240112305,
      "learning_rate": 0.0009315215102771411,
      "loss": 0.14814843237400055,
      "step": 1792
    },
    {
      "epoch": 0.9481481481481482,
      "grad_norm": 0.09348400682210922,
      "learning_rate": 0.0009175373450421618,
      "loss": 0.15098626911640167,
      "step": 1920
    },
    {
      "epoch": 1.011358024691358,
      "grad_norm": 0.03807132691144943,
      "learning_rate": 0.0009023808259343743,
      "loss": 0.14861960709095,
      "step": 2048
    },
    {
      "epoch": 1.011358024691358,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03768022817264243,
      "eval_loss": 0.14497914565045658,
      "eval_mse_loss": 0.14497914565045658,
      "step": 2048
    },
    {
      "epoch": 1.011358024691358,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03768022817264243,
      "eval_loss": 0.14497914565045658,
      "eval_mse_loss": 0.14497914565045658,
      "eval_runtime": 17.3913,
      "eval_samples_per_second": 150.535,
      "eval_steps_per_second": 2.357,
      "step": 2048
    },
    {
      "epoch": 1.074567901234568,
      "grad_norm": 0.047278061509132385,
      "learning_rate": 0.0008860945091564399,
      "loss": 0.14657916128635406,
      "step": 2176
    },
    {
      "epoch": 1.1377777777777778,
      "grad_norm": 0.04802122339606285,
      "learning_rate": 0.0008687241231366662,
      "loss": 0.1490355134010315,
      "step": 2304
    },
    {
      "epoch": 1.2009876543209876,
      "grad_norm": 0.047573771327733994,
      "learning_rate": 0.0008503184401335448,
      "loss": 0.15079016983509064,
      "step": 2432
    },
    {
      "epoch": 1.2641975308641975,
      "grad_norm": 0.051725756376981735,
      "learning_rate": 0.0008309291392938795,
      "loss": 0.15022365748882294,
      "step": 2560
    },
    {
      "epoch": 1.3274074074074074,
      "grad_norm": 0.027799006551504135,
      "learning_rate": 0.0008106106615490032,
      "loss": 0.1516639143228531,
      "step": 2688
    },
    {
      "epoch": 1.3906172839506172,
      "grad_norm": 0.07822522521018982,
      "learning_rate": 0.0007894200567565075,
      "loss": 0.15127751231193542,
      "step": 2816
    },
    {
      "epoch": 1.453827160493827,
      "grad_norm": 0.050239723175764084,
      "learning_rate": 0.0007674168235166747,
      "loss": 0.14863863587379456,
      "step": 2944
    },
    {
      "epoch": 1.5170370370370372,
      "grad_norm": 0.07048199325799942,
      "learning_rate": 0.000744662742113375,
      "loss": 0.14983540773391724,
      "step": 3072
    },
    {
      "epoch": 1.5170370370370372,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03912174238300905,
      "eval_loss": 0.151152161926758,
      "eval_mse_loss": 0.151152161926758,
      "step": 3072
    },
    {
      "epoch": 1.5170370370370372,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03912174238300905,
      "eval_loss": 0.151152161926758,
      "eval_mse_loss": 0.151152161926758,
      "eval_runtime": 17.2135,
      "eval_samples_per_second": 152.09,
      "eval_steps_per_second": 2.382,
      "step": 3072
    },
    {
      "epoch": 1.5802469135802468,
      "grad_norm": 0.03784182295203209,
      "learning_rate": 0.0007212217010484917,
      "loss": 0.1514694094657898,
      "step": 3200
    },
    {
      "epoch": 1.643456790123457,
      "grad_norm": 0.02600058726966381,
      "learning_rate": 0.0006971595176569332,
      "loss": 0.14975027740001678,
      "step": 3328
    },
    {
      "epoch": 1.7066666666666666,
      "grad_norm": 0.0323154516518116,
      "learning_rate": 0.0006725437533058971,
      "loss": 0.1495063155889511,
      "step": 3456
    },
    {
      "epoch": 1.7698765432098766,
      "grad_norm": 0.032654162496328354,
      "learning_rate": 0.0006474435236972767,
      "loss": 0.15044330060482025,
      "step": 3584
    },
    {
      "epoch": 1.8330864197530863,
      "grad_norm": 0.054954059422016144,
      "learning_rate": 0.0006219293048058301,
      "loss": 0.14830181002616882,
      "step": 3712
    },
    {
      "epoch": 1.8962962962962964,
      "grad_norm": 0.03366395831108093,
      "learning_rate": 0.0005960727349980042,
      "loss": 0.14830124378204346,
      "step": 3840
    },
    {
      "epoch": 1.959506172839506,
      "grad_norm": 0.0262282807379961,
      "learning_rate": 0.0005699464138870106,
      "loss": 0.14904989302158356,
      "step": 3968
    },
    {
      "epoch": 2.022716049382716,
      "grad_norm": 0.029652466997504234,
      "learning_rate": 0.0005436236984889356,
      "loss": 0.1490490734577179,
      "step": 4096
    },
    {
      "epoch": 2.022716049382716,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03822143793832965,
      "eval_loss": 0.14761792813859334,
      "eval_mse_loss": 0.14761792813859334,
      "step": 4096
    },
    {
      "epoch": 2.022716049382716,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03822143793832965,
      "eval_loss": 0.14761792813859334,
      "eval_mse_loss": 0.14761792813859334,
      "eval_runtime": 17.5759,
      "eval_samples_per_second": 148.954,
      "eval_steps_per_second": 2.333,
      "step": 4096
    },
    {
      "epoch": 2.0859259259259257,
      "grad_norm": 0.02397269941866398,
      "learning_rate": 0.0005171784972522235,
      "loss": 0.15046730637550354,
      "step": 4224
    },
    {
      "epoch": 2.149135802469136,
      "grad_norm": 0.03038075752556324,
      "learning_rate": 0.0004906850625388614,
      "loss": 0.1505787968635559,
      "step": 4352
    },
    {
      "epoch": 2.212345679012346,
      "grad_norm": 0.030504263937473297,
      "learning_rate": 0.0004642177821399269,
      "loss": 0.14973199367523193,
      "step": 4480
    },
    {
      "epoch": 2.2755555555555556,
      "grad_norm": 0.02406897582113743,
      "learning_rate": 0.00043785097041088293,
      "loss": 0.1480320692062378,
      "step": 4608
    },
    {
      "epoch": 2.338765432098765,
      "grad_norm": 0.025209980085492134,
      "learning_rate": 0.00041165865961306135,
      "loss": 0.15078303217887878,
      "step": 4736
    },
    {
      "epoch": 2.4019753086419753,
      "grad_norm": 0.021273162215948105,
      "learning_rate": 0.000385714392047208,
      "loss": 0.14956723153591156,
      "step": 4864
    },
    {
      "epoch": 2.4651851851851854,
      "grad_norm": 0.019383525475859642,
      "learning_rate": 0.00036009101356272814,
      "loss": 0.1489405333995819,
      "step": 4992
    },
    {
      "epoch": 2.528395061728395,
      "grad_norm": 0.020899692550301552,
      "learning_rate": 0.00033486046902241664,
      "loss": 0.14915253221988678,
      "step": 5120
    },
    {
      "epoch": 2.528395061728395,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03704260857548655,
      "eval_loss": 0.14219033936174905,
      "eval_mse_loss": 0.14219033936174905,
      "step": 5120
    },
    {
      "epoch": 2.528395061728395,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.03704260857548655,
      "eval_loss": 0.14219033936174905,
      "eval_mse_loss": 0.14219033936174905,
      "eval_runtime": 17.4466,
      "eval_samples_per_second": 150.058,
      "eval_steps_per_second": 2.35,
      "step": 5120
    },
    {
      "epoch": 2.591604938271605,
      "grad_norm": 0.02908194251358509,
      "learning_rate": 0.00031009360029696107,
      "loss": 0.14911262691020966,
      "step": 5248
    },
    {
      "epoch": 2.6548148148148147,
      "grad_norm": 0.06822603195905685,
      "learning_rate": 0.00028585994735640823,
      "loss": 0.1485205888748169,
      "step": 5376
    },
    {
      "epoch": 2.718024691358025,
      "grad_norm": 0.02113441936671734,
      "learning_rate": 0.0002622275530170825,
      "loss": 0.14886020123958588,
      "step": 5504
    },
    {
      "epoch": 2.7812345679012345,
      "grad_norm": 0.026887331157922745,
      "learning_rate": 0.0002392627718921852,
      "loss": 0.14821897447109222,
      "step": 5632
    },
    {
      "epoch": 2.8444444444444446,
      "grad_norm": 0.029895195737481117,
      "learning_rate": 0.00021703008408250187,
      "loss": 0.14855998754501343,
      "step": 5760
    },
    {
      "epoch": 2.907654320987654,
      "grad_norm": 0.017637664452195168,
      "learning_rate": 0.00019559191413033017,
      "loss": 0.14899146556854248,
      "step": 5888
    },
    {
      "epoch": 2.9708641975308643,
      "grad_norm": 0.035336632281541824,
      "learning_rate": 0.00017500845574497032,
      "loss": 0.14816507697105408,
      "step": 6016
    },
    {
      "epoch": 3.034074074074074,
      "grad_norm": 0.019196584820747375,
      "learning_rate": 0.00015533750279190617,
      "loss": 0.14883184432983398,
      "step": 6144
    },
    {
      "epoch": 3.034074074074074,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.038591444583200823,
      "eval_loss": 0.1482949287789624,
      "eval_mse_loss": 0.1482949287789624,
      "step": 6144
    },
    {
      "epoch": 3.034074074074074,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.038591444583200823,
      "eval_loss": 0.1482949287789624,
      "eval_mse_loss": 0.1482949287789624,
      "eval_runtime": 17.2605,
      "eval_samples_per_second": 151.676,
      "eval_steps_per_second": 2.375,
      "step": 6144
    },
    {
      "epoch": 3.097283950617284,
      "grad_norm": 0.02671050652861595,
      "learning_rate": 0.000136634287020226,
      "loss": 0.14740729331970215,
      "step": 6272
    },
    {
      "epoch": 3.1604938271604937,
      "grad_norm": 0.026964925229549408,
      "learning_rate": 0.00011895132298390743,
      "loss": 0.14907075464725494,
      "step": 6400
    },
    {
      "epoch": 3.2237037037037037,
      "grad_norm": 0.03113582730293274,
      "learning_rate": 0.00010233826059239426,
      "loss": 0.15129905939102173,
      "step": 6528
    },
    {
      "epoch": 3.286913580246914,
      "grad_norm": 0.016162540763616562,
      "learning_rate": 8.68417457044704e-05,
      "loss": 0.1497078388929367,
      "step": 6656
    },
    {
      "epoch": 3.3501234567901235,
      "grad_norm": 0.03715677186846733,
      "learning_rate": 7.250528915685422e-05,
      "loss": 0.15141023695468903,
      "step": 6784
    },
    {
      "epoch": 3.413333333333333,
      "grad_norm": 0.023826338350772858,
      "learning_rate": 5.93691445952505e-05,
      "loss": 0.14596687257289886,
      "step": 6912
    },
    {
      "epoch": 3.476543209876543,
      "grad_norm": 0.015810539945960045,
      "learning_rate": 4.7470195450886376e-05,
      "loss": 0.1492277979850769,
      "step": 7040
    },
    {
      "epoch": 3.5397530864197533,
      "grad_norm": 0.01591351442039013,
      "learning_rate": 3.684185137987378e-05,
      "loss": 0.1490740180015564,
      "step": 7168
    },
    {
      "epoch": 3.5397530864197533,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.038417009682190126,
      "eval_loss": 0.14850174581132283,
      "eval_mse_loss": 0.14850174581132283,
      "step": 7168
    },
    {
      "epoch": 3.5397530864197533,
      "eval_bleu": 0.0,
      "eval_cos_loss": 0.038417009682190126,
      "eval_loss": 0.14850174581132283,
      "eval_mse_loss": 0.14850174581132283,
      "eval_runtime": 17.2573,
      "eval_samples_per_second": 151.704,
      "eval_steps_per_second": 2.376,
      "step": 7168
    },
    {
      "epoch": 3.602962962962963,
      "grad_norm": 0.01680172048509121,
      "learning_rate": 2.751395445617594e-05,
      "loss": 0.1489180326461792,
      "step": 7296
    },
    {
      "epoch": 3.6661728395061726,
      "grad_norm": 0.018111728131771088,
      "learning_rate": 1.9512695381567302e-05,
      "loss": 0.1487143635749817,
      "step": 7424
    },
    {
      "epoch": 3.7293827160493827,
      "grad_norm": 0.03240951895713806,
      "learning_rate": 1.2860539947850303e-05,
      "loss": 0.1489432007074356,
      "step": 7552
    },
    {
      "epoch": 3.7925925925925927,
      "grad_norm": 0.025667285546660423,
      "learning_rate": 7.5761659578078565e-06,
      "loss": 0.15040472149848938,
      "step": 7680
    },
    {
      "epoch": 3.8558024691358024,
      "grad_norm": 0.01822933368384838,
      "learning_rate": 3.674410782003812e-06,
      "loss": 0.15010671317577362,
      "step": 7808
    },
    {
      "epoch": 3.9190123456790125,
      "grad_norm": 0.023645251989364624,
      "learning_rate": 1.1662296986795217e-06,
      "loss": 0.1507875919342041,
      "step": 7936
    },
    {
      "epoch": 3.982222222222222,
      "grad_norm": 0.02925770729780197,
      "learning_rate": 5.866513372004834e-08,
      "loss": 0.1489885449409485,
      "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
}