{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 715,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 11.76454607734096,
      "learning_rate": 3.4615384615384617e-06,
      "loss": 0.4865,
      "step": 10
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 14.968705460305152,
      "learning_rate": 4.99743403167199e-06,
      "loss": 0.359,
      "step": 20
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 8.988922429392048,
      "learning_rate": 4.981772185245135e-06,
      "loss": 0.3123,
      "step": 30
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 7.483069905570552,
      "learning_rate": 4.9519632010080765e-06,
      "loss": 0.2929,
      "step": 40
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 7.547133087976635,
      "learning_rate": 4.9081770026187915e-06,
      "loss": 0.2738,
      "step": 50
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 6.164825995820736,
      "learning_rate": 4.850663189692619e-06,
      "loss": 0.2622,
      "step": 60
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 10.82949040328208,
      "learning_rate": 4.779749614980225e-06,
      "loss": 0.2545,
      "step": 70
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 9.291675578653022,
      "learning_rate": 4.695840515470016e-06,
      "loss": 0.2443,
      "step": 80
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 4.483428430129757,
      "learning_rate": 4.5994142080684956e-06,
      "loss": 0.24,
      "step": 90
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 9.467430683733449,
      "learning_rate": 4.491020362994168e-06,
      "loss": 0.2357,
      "step": 100
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 8.360804191198046,
      "learning_rate": 4.3712768704277535e-06,
      "loss": 0.2423,
      "step": 110
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 9.376024777417225,
      "learning_rate": 4.240866318280133e-06,
      "loss": 0.2426,
      "step": 120
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 42.513122163358034,
      "learning_rate": 4.10053210115622e-06,
      "loss": 0.2431,
      "step": 130
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 10.60774073899308,
      "learning_rate": 3.9510741826952835e-06,
      "loss": 0.2296,
      "step": 140
    },
    {
      "epoch": 1.0491228070175438,
      "grad_norm": 7.049299408091297,
      "learning_rate": 3.793344535444142e-06,
      "loss": 0.2265,
      "step": 150
    },
    {
      "epoch": 1.119298245614035,
      "grad_norm": 66.59836296802126,
      "learning_rate": 3.6282422842578846e-06,
      "loss": 0.2229,
      "step": 160
    },
    {
      "epoch": 1.1894736842105262,
      "grad_norm": 31.009452424712197,
      "learning_rate": 3.4567085809127247e-06,
      "loss": 0.2255,
      "step": 170
    },
    {
      "epoch": 1.2596491228070175,
      "grad_norm": 4.707428577348247,
      "learning_rate": 3.2797212391478727e-06,
      "loss": 0.2208,
      "step": 180
    },
    {
      "epoch": 1.329824561403509,
      "grad_norm": 6.260813622561182,
      "learning_rate": 3.0982891607188948e-06,
      "loss": 0.226,
      "step": 190
    },
    {
      "epoch": 1.4,
      "grad_norm": 6.891400739921387,
      "learning_rate": 2.9134465842364034e-06,
      "loss": 0.2106,
      "step": 200
    },
    {
      "epoch": 1.4701754385964914,
      "grad_norm": 6.707681637517704,
      "learning_rate": 2.726247189574095e-06,
      "loss": 0.2164,
      "step": 210
    },
    {
      "epoch": 1.5403508771929824,
      "grad_norm": 14.86073483730446,
      "learning_rate": 2.537758091453465e-06,
      "loss": 0.2192,
      "step": 220
    },
    {
      "epoch": 1.6105263157894738,
      "grad_norm": 4.385122732812432,
      "learning_rate": 2.349053756444285e-06,
      "loss": 0.2189,
      "step": 230
    },
    {
      "epoch": 1.6807017543859648,
      "grad_norm": 7.004818522116804,
      "learning_rate": 2.1612098780564715e-06,
      "loss": 0.2126,
      "step": 240
    },
    {
      "epoch": 1.7508771929824563,
      "grad_norm": 6.590975514106617,
      "learning_rate": 1.9752972448378817e-06,
      "loss": 0.2145,
      "step": 250
    },
    {
      "epoch": 1.8210526315789473,
      "grad_norm": 9.753821755878821,
      "learning_rate": 1.7923756364324494e-06,
      "loss": 0.2159,
      "step": 260
    },
    {
      "epoch": 1.8912280701754387,
      "grad_norm": 5.325787513700657,
      "learning_rate": 1.613487782393661e-06,
      "loss": 0.2162,
      "step": 270
    },
    {
      "epoch": 1.9614035087719297,
      "grad_norm": 4.491557835089463,
      "learning_rate": 1.4396534181906726e-06,
      "loss": 0.2095,
      "step": 280
    },
    {
      "epoch": 2.0280701754385966,
      "grad_norm": 12.559781170205573,
      "learning_rate": 1.2718634722903073e-06,
      "loss": 0.2099,
      "step": 290
    },
    {
      "epoch": 2.0982456140350876,
      "grad_norm": 5.079848672597547,
      "learning_rate": 1.1110744174509952e-06,
      "loss": 0.209,
      "step": 300
    },
    {
      "epoch": 2.168421052631579,
      "grad_norm": 10.17251965492991,
      "learning_rate": 9.582028184286423e-07,
      "loss": 0.2055,
      "step": 310
    },
    {
      "epoch": 2.23859649122807,
      "grad_norm": 13.627772602845546,
      "learning_rate": 8.141201071747784e-07,
      "loss": 0.203,
      "step": 320
    },
    {
      "epoch": 2.3087719298245615,
      "grad_norm": 8.355002030998532,
      "learning_rate": 6.796476153105294e-07,
      "loss": 0.2091,
      "step": 330
    },
    {
      "epoch": 2.3789473684210525,
      "grad_norm": 4.530969715092393,
      "learning_rate": 5.555518921934047e-07,
      "loss": 0.2084,
      "step": 340
    },
    {
      "epoch": 2.449122807017544,
      "grad_norm": 6.272704466018619,
      "learning_rate": 4.4254033526585917e-07,
      "loss": 0.2118,
      "step": 350
    },
    {
      "epoch": 2.519298245614035,
      "grad_norm": 14.525483822981093,
      "learning_rate": 3.412571575944679e-07,
      "loss": 0.2087,
      "step": 360
    },
    {
      "epoch": 2.5894736842105264,
      "grad_norm": 8.530705356489067,
      "learning_rate": 2.522797155864354e-07,
      "loss": 0.2053,
      "step": 370
    },
    {
      "epoch": 2.659649122807018,
      "grad_norm": 21.754664451542073,
      "learning_rate": 1.7611521781697644e-07,
      "loss": 0.2015,
      "step": 380
    },
    {
      "epoch": 2.729824561403509,
      "grad_norm": 5.016185711096282,
      "learning_rate": 1.1319783372863601e-07,
      "loss": 0.2029,
      "step": 390
    },
    {
      "epoch": 2.8,
      "grad_norm": 5.315305503621668,
      "learning_rate": 6.388621868416101e-08,
      "loss": 0.2019,
      "step": 400
    },
    {
      "epoch": 2.8701754385964913,
      "grad_norm": 8.076616074629916,
      "learning_rate": 2.8461469481164682e-08,
      "loss": 0.209,
      "step": 410
    },
    {
      "epoch": 2.9403508771929827,
      "grad_norm": 8.636591588732433,
      "learning_rate": 7.125521982997152e-09,
      "loss": 0.2039,
      "step": 420
    },
    {
      "epoch": 3.007017543859649,
      "grad_norm": 7.41019572219896,
      "learning_rate": 0.0,
      "loss": 0.1973,
      "step": 430
    },
    {
      "epoch": 3.07719298245614,
      "grad_norm": 5.298408220149024,
      "learning_rate": 1.7145346705573441e-06,
      "loss": 0.2047,
      "step": 440
    },
    {
      "epoch": 3.1473684210526316,
      "grad_norm": 7.552251123214438,
      "learning_rate": 1.6077844460203207e-06,
      "loss": 0.2043,
      "step": 450
    },
    {
      "epoch": 3.2175438596491226,
      "grad_norm": 5.362701605328588,
      "learning_rate": 1.502867501009409e-06,
      "loss": 0.2069,
      "step": 460
    },
    {
      "epoch": 3.287719298245614,
      "grad_norm": 4.441253980224051,
      "learning_rate": 1.3999994135722822e-06,
      "loss": 0.2041,
      "step": 470
    },
    {
      "epoch": 3.3578947368421055,
      "grad_norm": 4.970304301865314,
      "learning_rate": 1.2993915518673966e-06,
      "loss": 0.2082,
      "step": 480
    },
    {
      "epoch": 3.4280701754385965,
      "grad_norm": 23.210383061214984,
      "learning_rate": 1.20125063985535e-06,
      "loss": 0.214,
      "step": 490
    },
    {
      "epoch": 3.4982456140350875,
      "grad_norm": 4.220133166172592,
      "learning_rate": 1.105778332532927e-06,
      "loss": 0.2005,
      "step": 500
    },
    {
      "epoch": 3.568421052631579,
      "grad_norm": 4.116716856261486,
      "learning_rate": 1.0131708015825668e-06,
      "loss": 0.2105,
      "step": 510
    },
    {
      "epoch": 3.6385964912280704,
      "grad_norm": 7.3207052924218665,
      "learning_rate": 9.236183322886946e-07,
      "loss": 0.2033,
      "step": 520
    },
    {
      "epoch": 3.7087719298245614,
      "grad_norm": 3.852808758513771,
      "learning_rate": 8.373049325491036e-07,
      "loss": 0.1985,
      "step": 530
    },
    {
      "epoch": 3.7789473684210524,
      "grad_norm": 15.250294356744208,
      "learning_rate": 7.544079547848183e-07,
      "loss": 0.1964,
      "step": 540
    },
    {
      "epoch": 3.849122807017544,
      "grad_norm": 3.0509014746790166,
      "learning_rate": 6.750977315252732e-07,
      "loss": 0.2008,
      "step": 550
    },
    {
      "epoch": 3.9192982456140353,
      "grad_norm": 4.687758452399206,
      "learning_rate": 5.995372254176302e-07,
      "loss": 0.2097,
      "step": 560
    },
    {
      "epoch": 3.9894736842105263,
      "grad_norm": 9.721911452314314,
      "learning_rate": 5.278816943793382e-07,
      "loss": 0.201,
      "step": 570
    },
    {
      "epoch": 4.056140350877193,
      "grad_norm": 2.403043490486729,
      "learning_rate": 4.602783725820076e-07,
      "loss": 0.1967,
      "step": 580
    },
    {
      "epoch": 4.126315789473685,
      "grad_norm": 8.290723653754437,
      "learning_rate": 3.9686616792204677e-07,
      "loss": 0.1979,
      "step": 590
    },
    {
      "epoch": 4.196491228070175,
      "grad_norm": 5.444718615266718,
      "learning_rate": 3.3777537659973184e-07,
      "loss": 0.2002,
      "step": 600
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 6.721894904138142,
      "learning_rate": 2.831274153931357e-07,
      "loss": 0.2005,
      "step": 610
    },
    {
      "epoch": 4.336842105263158,
      "grad_norm": 15.697675859013298,
      "learning_rate": 2.330345721770641e-07,
      "loss": 0.1974,
      "step": 620
    },
    {
      "epoch": 4.4070175438596495,
      "grad_norm": 10.41027198048078,
      "learning_rate": 1.875997751995881e-07,
      "loss": 0.1983,
      "step": 630
    },
    {
      "epoch": 4.47719298245614,
      "grad_norm": 5.6755064919838345,
      "learning_rate": 1.4691638159027105e-07,
      "loss": 0.1941,
      "step": 640
    },
    {
      "epoch": 4.5473684210526315,
      "grad_norm": 15.635280214115102,
      "learning_rate": 1.1106798553464804e-07,
      "loss": 0.199,
      "step": 650
    },
    {
      "epoch": 4.617543859649123,
      "grad_norm": 16.997926750117426,
      "learning_rate": 8.012824650910938e-08,
      "loss": 0.2002,
      "step": 660
    },
    {
      "epoch": 4.687719298245614,
      "grad_norm": 11.260397619043863,
      "learning_rate": 5.416073792912091e-08,
      "loss": 0.1999,
      "step": 670
    },
    {
      "epoch": 4.757894736842105,
      "grad_norm": 4.712188848713119,
      "learning_rate": 3.321881652177783e-08,
      "loss": 0.198,
      "step": 680
    },
    {
      "epoch": 4.828070175438596,
      "grad_norm": 6.993435000049924,
      "learning_rate": 1.7345512691092083e-08,
      "loss": 0.204,
      "step": 690
    },
    {
      "epoch": 4.898245614035088,
      "grad_norm": 5.7468977622367055,
      "learning_rate": 6.57344210128863e-09,
      "loss": 0.2001,
      "step": 700
    },
    {
      "epoch": 4.968421052631579,
      "grad_norm": 5.717757397470271,
      "learning_rate": 9.247386597824315e-10,
      "loss": 0.1992,
      "step": 710
    },
    {
      "epoch": 5.0,
      "step": 715,
      "total_flos": 2064156363063296.0,
      "train_loss": 0.08062235821377148,
      "train_runtime": 7241.919,
      "train_samples_per_second": 6.295,
      "train_steps_per_second": 0.099
    }
  ],
  "logging_steps": 10,
  "max_steps": 715,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 2064156363063296.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}