{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 313,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "alignment_loss": 0.956536,
      "epoch": 0.032,
      "grad_norm": 8938.3779296875,
      "learning_rate": 9e-06,
      "loss": 537.496,
      "mean_token_accuracy": 0.6349152714014054,
      "num_tokens": 684434.0,
      "step": 10,
      "teacher_ce_loss": 67.311574
    },
    {
      "alignment_loss": 0.953701,
      "epoch": 0.064,
      "grad_norm": 2418.03369140625,
      "learning_rate": 9.702970297029704e-06,
      "loss": 148.2146,
      "mean_token_accuracy": 0.7730402827262879,
      "num_tokens": 1350616.0,
      "step": 20,
      "teacher_ce_loss": 19.467055
    },
    {
      "alignment_loss": 0.953897,
      "epoch": 0.096,
      "grad_norm": 2525.4296875,
      "learning_rate": 9.372937293729373e-06,
      "loss": 106.736,
      "mean_token_accuracy": 0.7940315298736096,
      "num_tokens": 2048862.0,
      "step": 30,
      "teacher_ce_loss": 13.013605
    },
    {
      "alignment_loss": 0.947818,
      "epoch": 0.128,
      "grad_norm": 2252.9521484375,
      "learning_rate": 9.042904290429043e-06,
      "loss": 79.3102,
      "mean_token_accuracy": 0.8635720252990723,
      "num_tokens": 2752423.0,
      "step": 40,
      "teacher_ce_loss": 9.476292
    },
    {
      "alignment_loss": 0.93975,
      "epoch": 0.16,
      "grad_norm": 941.5345458984375,
      "learning_rate": 8.712871287128714e-06,
      "loss": 57.8999,
      "mean_token_accuracy": 0.8709644131362438,
      "num_tokens": 3415991.0,
      "step": 50,
      "teacher_ce_loss": 7.637903
    },
    {
      "alignment_loss": 0.93242,
      "epoch": 0.192,
      "grad_norm": 494.6402587890625,
      "learning_rate": 8.382838283828383e-06,
      "loss": 49.9108,
      "mean_token_accuracy": 0.8711144886910915,
      "num_tokens": 4092597.0,
      "step": 60,
      "teacher_ce_loss": 5.167176
    },
    {
      "alignment_loss": 0.927774,
      "epoch": 0.224,
      "grad_norm": 448.8665466308594,
      "learning_rate": 8.052805280528053e-06,
      "loss": 48.3418,
      "mean_token_accuracy": 0.8714926809072494,
      "num_tokens": 4782962.0,
      "step": 70,
      "teacher_ce_loss": 6.822194
    },
    {
      "alignment_loss": 0.922368,
      "epoch": 0.256,
      "grad_norm": 512.7625122070312,
      "learning_rate": 7.722772277227724e-06,
      "loss": 50.2869,
      "mean_token_accuracy": 0.8667933315038681,
      "num_tokens": 5463890.0,
      "step": 80,
      "teacher_ce_loss": 6.749592
    },
    {
      "alignment_loss": 0.914546,
      "epoch": 0.288,
      "grad_norm": 405.1988220214844,
      "learning_rate": 7.392739273927393e-06,
      "loss": 49.6484,
      "mean_token_accuracy": 0.8680442161858082,
      "num_tokens": 6146518.0,
      "step": 90,
      "teacher_ce_loss": 7.311445
    },
    {
      "alignment_loss": 0.910286,
      "epoch": 0.32,
      "grad_norm": 389.6947937011719,
      "learning_rate": 7.062706270627063e-06,
      "loss": 50.4636,
      "mean_token_accuracy": 0.8658453419804573,
      "num_tokens": 6826396.0,
      "step": 100,
      "teacher_ce_loss": 5.171713
    },
    {
      "alignment_loss": 0.902513,
      "epoch": 0.352,
      "grad_norm": 429.2699890136719,
      "learning_rate": 6.732673267326733e-06,
      "loss": 47.7155,
      "mean_token_accuracy": 0.8693347111344337,
      "num_tokens": 7517684.0,
      "step": 110,
      "teacher_ce_loss": 5.650189
    },
    {
      "alignment_loss": 0.896312,
      "epoch": 0.384,
      "grad_norm": 430.8398742675781,
      "learning_rate": 6.402640264026403e-06,
      "loss": 47.995,
      "mean_token_accuracy": 0.8666416265070438,
      "num_tokens": 8188946.0,
      "step": 120,
      "teacher_ce_loss": 7.503155
    },
    {
      "alignment_loss": 0.893979,
      "epoch": 0.416,
      "grad_norm": 791.207275390625,
      "learning_rate": 6.072607260726073e-06,
      "loss": 46.9929,
      "mean_token_accuracy": 0.873030910640955,
      "num_tokens": 8874685.0,
      "step": 130,
      "teacher_ce_loss": 5.863695
    },
    {
      "alignment_loss": 0.892291,
      "epoch": 0.448,
      "grad_norm": 438.9299011230469,
      "learning_rate": 5.7425742574257425e-06,
      "loss": 46.5332,
      "mean_token_accuracy": 0.8730357453227043,
      "num_tokens": 9566920.0,
      "step": 140,
      "teacher_ce_loss": 4.956036
    },
    {
      "alignment_loss": 0.887302,
      "epoch": 0.48,
      "grad_norm": 455.186767578125,
      "learning_rate": 5.412541254125413e-06,
      "loss": 47.4667,
      "mean_token_accuracy": 0.8698927395045757,
      "num_tokens": 10249181.0,
      "step": 150,
      "teacher_ce_loss": 5.094092
    },
    {
      "alignment_loss": 0.887032,
      "epoch": 0.512,
      "grad_norm": 398.2301330566406,
      "learning_rate": 5.082508250825083e-06,
      "loss": 44.8594,
      "mean_token_accuracy": 0.8733536802232266,
      "num_tokens": 10942070.0,
      "step": 160,
      "teacher_ce_loss": 6.080262
    },
    {
      "alignment_loss": 0.885581,
      "epoch": 0.544,
      "grad_norm": 391.7760314941406,
      "learning_rate": 4.7524752475247525e-06,
      "loss": 46.027,
      "mean_token_accuracy": 0.8712647803127765,
      "num_tokens": 11634781.0,
      "step": 170,
      "teacher_ce_loss": 5.655279
    },
    {
      "alignment_loss": 0.883271,
      "epoch": 0.576,
      "grad_norm": 411.50592041015625,
      "learning_rate": 4.422442244224423e-06,
      "loss": 45.4771,
      "mean_token_accuracy": 0.8760336212813854,
      "num_tokens": 12331934.0,
      "step": 180,
      "teacher_ce_loss": 5.976733
    },
    {
      "alignment_loss": 0.879363,
      "epoch": 0.608,
      "grad_norm": 406.4468078613281,
      "learning_rate": 4.092409240924093e-06,
      "loss": 43.7762,
      "mean_token_accuracy": 0.8771503999829292,
      "num_tokens": 13004167.0,
      "step": 190,
      "teacher_ce_loss": 6.10556
    },
    {
      "alignment_loss": 0.873679,
      "epoch": 0.64,
      "grad_norm": 422.99945068359375,
      "learning_rate": 3.7623762376237625e-06,
      "loss": 44.5788,
      "mean_token_accuracy": 0.8752013951539993,
      "num_tokens": 13689914.0,
      "step": 200,
      "teacher_ce_loss": 5.429278
    },
    {
      "alignment_loss": 0.863939,
      "epoch": 0.672,
      "grad_norm": 415.8448486328125,
      "learning_rate": 3.4323432343234324e-06,
      "loss": 42.7953,
      "mean_token_accuracy": 0.8772109493613243,
      "num_tokens": 14378852.0,
      "step": 210,
      "teacher_ce_loss": 4.237569
    },
    {
      "alignment_loss": 0.862601,
      "epoch": 0.704,
      "grad_norm": 368.5125732421875,
      "learning_rate": 3.1023102310231023e-06,
      "loss": 42.6845,
      "mean_token_accuracy": 0.8811097137629986,
      "num_tokens": 15067474.0,
      "step": 220,
      "teacher_ce_loss": 5.223768
    },
    {
      "alignment_loss": 0.856529,
      "epoch": 0.736,
      "grad_norm": 400.0111083984375,
      "learning_rate": 2.7722772277227726e-06,
      "loss": 43.299,
      "mean_token_accuracy": 0.8764385782182217,
      "num_tokens": 15767042.0,
      "step": 230,
      "teacher_ce_loss": 5.99914
    },
    {
      "alignment_loss": 0.849113,
      "epoch": 0.768,
      "grad_norm": 412.9610900878906,
      "learning_rate": 2.4422442244224424e-06,
      "loss": 44.0378,
      "mean_token_accuracy": 0.8751519277691842,
      "num_tokens": 16463926.0,
      "step": 240,
      "teacher_ce_loss": 5.501055
    },
    {
      "alignment_loss": 0.850236,
      "epoch": 0.8,
      "grad_norm": 390.5102233886719,
      "learning_rate": 2.1122112211221123e-06,
      "loss": 44.4366,
      "mean_token_accuracy": 0.8718901112675667,
      "num_tokens": 17171085.0,
      "step": 250,
      "teacher_ce_loss": 6.008397
    },
    {
      "alignment_loss": 0.848085,
      "epoch": 0.832,
      "grad_norm": 373.6785583496094,
      "learning_rate": 1.7821782178217822e-06,
      "loss": 43.3382,
      "mean_token_accuracy": 0.8724160015583038,
      "num_tokens": 17873884.0,
      "step": 260,
      "teacher_ce_loss": 6.177311
    },
    {
      "alignment_loss": 0.842402,
      "epoch": 0.864,
      "grad_norm": 401.6341247558594,
      "learning_rate": 1.4521452145214523e-06,
      "loss": 42.9951,
      "mean_token_accuracy": 0.8800486765801907,
      "num_tokens": 18561026.0,
      "step": 270,
      "teacher_ce_loss": 5.229172
    },
    {
      "alignment_loss": 0.842598,
      "epoch": 0.896,
      "grad_norm": 348.2569885253906,
      "learning_rate": 1.1221122112211221e-06,
      "loss": 43.8781,
      "mean_token_accuracy": 0.8713223539292813,
      "num_tokens": 19268650.0,
      "step": 280,
      "teacher_ce_loss": 5.308041
    },
    {
      "alignment_loss": 0.840163,
      "epoch": 0.928,
      "grad_norm": 397.3896179199219,
      "learning_rate": 7.920792079207921e-07,
      "loss": 43.2348,
      "mean_token_accuracy": 0.8775620862841607,
      "num_tokens": 19979134.0,
      "step": 290,
      "teacher_ce_loss": 4.710304
    },
    {
      "alignment_loss": 0.842505,
      "epoch": 0.96,
      "grad_norm": 434.8416442871094,
      "learning_rate": 4.6204620462046204e-07,
      "loss": 42.5038,
      "mean_token_accuracy": 0.880195677280426,
      "num_tokens": 20671599.0,
      "step": 300,
      "teacher_ce_loss": 5.022283
    },
    {
      "alignment_loss": 0.84056,
      "epoch": 0.992,
      "grad_norm": 391.82354736328125,
      "learning_rate": 1.3201320132013202e-07,
      "loss": 42.0722,
      "mean_token_accuracy": 0.8764784179627896,
      "num_tokens": 21386663.0,
      "step": 310,
      "teacher_ce_loss": 5.886476
    }
  ],
  "logging_steps": 10,
  "max_steps": 313,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.002134616338858e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}