{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 720,
  "global_step": 17995,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13892747985551543,
      "grad_norm": 0.9573676586151123,
      "learning_rate": 0.0007907566916736131,
      "loss": 0.377,
      "step": 500
    },
    {
      "epoch": 0.2000555709919422,
      "eval_loss": 0.28939542174339294,
      "eval_runtime": 4.6518,
      "eval_samples_per_second": 107.484,
      "eval_steps_per_second": 6.879,
      "step": 720
    },
    {
      "epoch": 0.27785495971103086,
      "grad_norm": 0.5571001768112183,
      "learning_rate": 0.0007815133833472261,
      "loss": 0.2824,
      "step": 1000
    },
    {
      "epoch": 0.4001111419838844,
      "eval_loss": 0.2604603171348572,
      "eval_runtime": 4.3083,
      "eval_samples_per_second": 116.054,
      "eval_steps_per_second": 7.427,
      "step": 1440
    },
    {
      "epoch": 0.41678243956654626,
      "grad_norm": 0.6729081273078918,
      "learning_rate": 0.0007722515513568584,
      "loss": 0.2627,
      "step": 1500
    },
    {
      "epoch": 0.5557099194220617,
      "grad_norm": 0.5200651288032532,
      "learning_rate": 0.0007629897193664907,
      "loss": 0.2488,
      "step": 2000
    },
    {
      "epoch": 0.6001667129758266,
      "eval_loss": 0.24864767491817474,
      "eval_runtime": 3.8713,
      "eval_samples_per_second": 129.157,
      "eval_steps_per_second": 8.266,
      "step": 2160
    },
    {
      "epoch": 0.6946373992775771,
      "grad_norm": 0.6645380258560181,
      "learning_rate": 0.0007537464110401038,
      "loss": 0.2424,
      "step": 2500
    },
    {
      "epoch": 0.8002222839677688,
      "eval_loss": 0.2434273064136505,
      "eval_runtime": 4.2012,
      "eval_samples_per_second": 119.013,
      "eval_steps_per_second": 7.617,
      "step": 2880
    },
    {
      "epoch": 0.8335648791330925,
      "grad_norm": 0.8886778354644775,
      "learning_rate": 0.000744484579049736,
      "loss": 0.2332,
      "step": 3000
    },
    {
      "epoch": 0.972492358988608,
      "grad_norm": 0.5990277528762817,
      "learning_rate": 0.0007352227470593684,
      "loss": 0.2307,
      "step": 3500
    },
    {
      "epoch": 1.000277854959711,
      "eval_loss": 0.23320984840393066,
      "eval_runtime": 4.3777,
      "eval_samples_per_second": 114.216,
      "eval_steps_per_second": 7.31,
      "step": 3600
    },
    {
      "epoch": 1.1114198388441234,
      "grad_norm": 0.9014330506324768,
      "learning_rate": 0.0007259609150690007,
      "loss": 0.2171,
      "step": 4000
    },
    {
      "epoch": 1.2003334259516532,
      "eval_loss": 0.23165521025657654,
      "eval_runtime": 4.5533,
      "eval_samples_per_second": 109.81,
      "eval_steps_per_second": 7.028,
      "step": 4320
    },
    {
      "epoch": 1.2503473186996388,
      "grad_norm": 0.9070354104042053,
      "learning_rate": 0.000716699083078633,
      "loss": 0.2174,
      "step": 4500
    },
    {
      "epoch": 1.3892747985551543,
      "grad_norm": 0.9023842215538025,
      "learning_rate": 0.000707455774752246,
      "loss": 0.215,
      "step": 5000
    },
    {
      "epoch": 1.4003889969435954,
      "eval_loss": 0.21996016800403595,
      "eval_runtime": 4.2649,
      "eval_samples_per_second": 117.235,
      "eval_steps_per_second": 7.503,
      "step": 5040
    },
    {
      "epoch": 1.5282022784106695,
      "grad_norm": 0.8724926710128784,
      "learning_rate": 0.0006981939427618783,
      "loss": 0.2101,
      "step": 5500
    },
    {
      "epoch": 1.6004445679355377,
      "eval_loss": 0.21771642565727234,
      "eval_runtime": 4.6518,
      "eval_samples_per_second": 107.486,
      "eval_steps_per_second": 6.879,
      "step": 5760
    },
    {
      "epoch": 1.667129758266185,
      "grad_norm": 1.1229606866836548,
      "learning_rate": 0.0006889321107715106,
      "loss": 0.209,
      "step": 6000
    },
    {
      "epoch": 1.8005001389274797,
      "eval_loss": 0.2162657231092453,
      "eval_runtime": 3.9728,
      "eval_samples_per_second": 125.857,
      "eval_steps_per_second": 8.055,
      "step": 6480
    },
    {
      "epoch": 1.8060572381217006,
      "grad_norm": 0.562541127204895,
      "learning_rate": 0.000679670278781143,
      "loss": 0.2053,
      "step": 6500
    },
    {
      "epoch": 1.9449847179772157,
      "grad_norm": 0.9058884382247925,
      "learning_rate": 0.0006704084467907753,
      "loss": 0.2047,
      "step": 7000
    },
    {
      "epoch": 2.000555709919422,
      "eval_loss": 0.21751119196414948,
      "eval_runtime": 3.8866,
      "eval_samples_per_second": 128.646,
      "eval_steps_per_second": 8.233,
      "step": 7200
    },
    {
      "epoch": 2.0839121978327313,
      "grad_norm": 0.5113864541053772,
      "learning_rate": 0.0006611466148004076,
      "loss": 0.1972,
      "step": 7500
    },
    {
      "epoch": 2.2006112809113643,
      "eval_loss": 0.21081140637397766,
      "eval_runtime": 3.9021,
      "eval_samples_per_second": 128.136,
      "eval_steps_per_second": 8.201,
      "step": 7920
    },
    {
      "epoch": 2.222839677688247,
      "grad_norm": 0.6523454785346985,
      "learning_rate": 0.0006518847828100398,
      "loss": 0.1968,
      "step": 8000
    },
    {
      "epoch": 2.361767157543762,
      "grad_norm": 0.48720309138298035,
      "learning_rate": 0.0006426414744836529,
      "loss": 0.1934,
      "step": 8500
    },
    {
      "epoch": 2.4006668519033063,
      "eval_loss": 0.20452378690242767,
      "eval_runtime": 4.661,
      "eval_samples_per_second": 107.272,
      "eval_steps_per_second": 6.865,
      "step": 8640
    },
    {
      "epoch": 2.5006946373992776,
      "grad_norm": 0.6927788257598877,
      "learning_rate": 0.0006333796424932852,
      "loss": 0.1906,
      "step": 9000
    },
    {
      "epoch": 2.600722422895249,
      "eval_loss": 0.1991214007139206,
      "eval_runtime": 4.7026,
      "eval_samples_per_second": 106.324,
      "eval_steps_per_second": 6.805,
      "step": 9360
    },
    {
      "epoch": 2.639622117254793,
      "grad_norm": 0.6956183910369873,
      "learning_rate": 0.0006241178105029175,
      "loss": 0.1893,
      "step": 9500
    },
    {
      "epoch": 2.7785495971103087,
      "grad_norm": 0.40230458974838257,
      "learning_rate": 0.0006148559785125498,
      "loss": 0.1889,
      "step": 10000
    },
    {
      "epoch": 2.800777993887191,
      "eval_loss": 0.20080548524856567,
      "eval_runtime": 4.5769,
      "eval_samples_per_second": 109.245,
      "eval_steps_per_second": 6.992,
      "step": 10080
    },
    {
      "epoch": 2.917477076965824,
      "grad_norm": 1.1422420740127563,
      "learning_rate": 0.0006055941465221821,
      "loss": 0.1911,
      "step": 10500
    },
    {
      "epoch": 3.000833564879133,
      "eval_loss": 0.2006545066833496,
      "eval_runtime": 4.7461,
      "eval_samples_per_second": 105.351,
      "eval_steps_per_second": 6.742,
      "step": 10800
    },
    {
      "epoch": 3.0564045568213394,
      "grad_norm": 0.8359866142272949,
      "learning_rate": 0.0005963323145318144,
      "loss": 0.1858,
      "step": 11000
    },
    {
      "epoch": 3.1953320366768545,
      "grad_norm": 0.5189564824104309,
      "learning_rate": 0.0005870704825414468,
      "loss": 0.1789,
      "step": 11500
    },
    {
      "epoch": 3.2008891358710754,
      "eval_loss": 0.19731611013412476,
      "eval_runtime": 4.0407,
      "eval_samples_per_second": 123.742,
      "eval_steps_per_second": 7.919,
      "step": 11520
    },
    {
      "epoch": 3.33425951653237,
      "grad_norm": 0.578628659248352,
      "learning_rate": 0.0005778271742150597,
      "loss": 0.1799,
      "step": 12000
    },
    {
      "epoch": 3.4009447068630174,
      "eval_loss": 0.19467230141162872,
      "eval_runtime": 4.0481,
      "eval_samples_per_second": 123.514,
      "eval_steps_per_second": 7.905,
      "step": 12240
    },
    {
      "epoch": 3.4731869963878856,
      "grad_norm": 0.8292349576950073,
      "learning_rate": 0.0005685653422246921,
      "loss": 0.1769,
      "step": 12500
    },
    {
      "epoch": 3.6010002778549595,
      "eval_loss": 0.19331130385398865,
      "eval_runtime": 4.0814,
      "eval_samples_per_second": 122.508,
      "eval_steps_per_second": 7.841,
      "step": 12960
    },
    {
      "epoch": 3.612114476243401,
      "grad_norm": 0.6857467293739319,
      "learning_rate": 0.0005593035102343244,
      "loss": 0.1796,
      "step": 13000
    },
    {
      "epoch": 3.7510419560989163,
      "grad_norm": 0.5829126238822937,
      "learning_rate": 0.0005500416782439567,
      "loss": 0.176,
      "step": 13500
    },
    {
      "epoch": 3.801055848846902,
      "eval_loss": 0.18971021473407745,
      "eval_runtime": 4.5985,
      "eval_samples_per_second": 108.73,
      "eval_steps_per_second": 6.959,
      "step": 13680
    },
    {
      "epoch": 3.889969435954432,
      "grad_norm": 0.7102944254875183,
      "learning_rate": 0.0005407798462535889,
      "loss": 0.1753,
      "step": 14000
    },
    {
      "epoch": 4.001111419838844,
      "eval_loss": 0.18332147598266602,
      "eval_runtime": 4.569,
      "eval_samples_per_second": 109.434,
      "eval_steps_per_second": 7.004,
      "step": 14400
    },
    {
      "epoch": 4.0288969158099475,
      "grad_norm": 0.5539807081222534,
      "learning_rate": 0.000531536537927202,
      "loss": 0.17,
      "step": 14500
    },
    {
      "epoch": 4.167824395665463,
      "grad_norm": 0.6355165243148804,
      "learning_rate": 0.0005222747059368344,
      "loss": 0.1645,
      "step": 15000
    },
    {
      "epoch": 4.201166990830786,
      "eval_loss": 0.18329477310180664,
      "eval_runtime": 4.6958,
      "eval_samples_per_second": 106.477,
      "eval_steps_per_second": 6.815,
      "step": 15120
    },
    {
      "epoch": 4.306751875520978,
      "grad_norm": 0.5928105115890503,
      "learning_rate": 0.0005130128739464667,
      "loss": 0.1684,
      "step": 15500
    },
    {
      "epoch": 4.4012225618227285,
      "eval_loss": 0.1863529086112976,
      "eval_runtime": 3.9814,
      "eval_samples_per_second": 125.585,
      "eval_steps_per_second": 8.037,
      "step": 15840
    },
    {
      "epoch": 4.445679355376494,
      "grad_norm": 0.5573757886886597,
      "learning_rate": 0.000503751041956099,
      "loss": 0.1671,
      "step": 16000
    },
    {
      "epoch": 4.584606835232009,
      "grad_norm": 0.5758992433547974,
      "learning_rate": 0.0004944892099657312,
      "loss": 0.1626,
      "step": 16500
    },
    {
      "epoch": 4.601278132814671,
      "eval_loss": 0.18393146991729736,
      "eval_runtime": 3.9565,
      "eval_samples_per_second": 126.373,
      "eval_steps_per_second": 8.088,
      "step": 16560
    },
    {
      "epoch": 4.723534315087524,
      "grad_norm": 0.65135258436203,
      "learning_rate": 0.0004852459016393443,
      "loss": 0.1649,
      "step": 17000
    },
    {
      "epoch": 4.801333703806613,
      "eval_loss": 0.176134392619133,
      "eval_runtime": 4.7856,
      "eval_samples_per_second": 104.479,
      "eval_steps_per_second": 6.687,
      "step": 17280
    },
    {
      "epoch": 4.86246179494304,
      "grad_norm": 0.4339986741542816,
      "learning_rate": 0.0004759840696489766,
      "loss": 0.1643,
      "step": 17500
    }
  ],
  "logging_steps": 500,
  "max_steps": 43188,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.05936785054892e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}