{
  "best_global_step": 5445,
  "best_metric": 3.110565185546875,
  "best_model_checkpoint": null,
  "epoch": 17.171717171717173,
  "eval_steps": 500,
  "global_step": 8500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00202020202020202,
      "grad_norm": 27.219696044921875,
      "learning_rate": 0.0,
      "loss": 10.3707,
      "step": 1
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.5280556678771973,
      "learning_rate": 9.88e-05,
      "loss": 4.8958,
      "step": 495
    },
    {
      "epoch": 1.0,
      "eval_loss": 3.9314539432525635,
      "eval_runtime": 29.8878,
      "eval_samples_per_second": 1062.842,
      "eval_steps_per_second": 4.182,
      "step": 495
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.212111234664917,
      "learning_rate": 9.478678038379531e-05,
      "loss": 3.6499,
      "step": 990
    },
    {
      "epoch": 2.0,
      "eval_loss": 3.539745330810547,
      "eval_runtime": 29.254,
      "eval_samples_per_second": 1085.869,
      "eval_steps_per_second": 4.273,
      "step": 990
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.3252358436584473,
      "learning_rate": 8.950959488272922e-05,
      "loss": 3.3793,
      "step": 1485
    },
    {
      "epoch": 3.0,
      "eval_loss": 3.379368305206299,
      "eval_runtime": 30.2558,
      "eval_samples_per_second": 1049.913,
      "eval_steps_per_second": 4.131,
      "step": 1485
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.02847957611084,
      "learning_rate": 8.42324093816631e-05,
      "loss": 3.2264,
      "step": 1980
    },
    {
      "epoch": 4.0,
      "eval_loss": 3.2874038219451904,
      "eval_runtime": 30.0535,
      "eval_samples_per_second": 1056.98,
      "eval_steps_per_second": 4.159,
      "step": 1980
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.101438045501709,
      "learning_rate": 7.895522388059702e-05,
      "loss": 3.1179,
      "step": 2475
    },
    {
      "epoch": 5.0,
      "eval_loss": 3.2294254302978516,
      "eval_runtime": 27.6091,
      "eval_samples_per_second": 1150.562,
      "eval_steps_per_second": 4.527,
      "step": 2475
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.4466798305511475,
      "learning_rate": 7.367803837953093e-05,
      "loss": 3.0343,
      "step": 2970
    },
    {
      "epoch": 6.0,
      "eval_loss": 3.183612585067749,
      "eval_runtime": 29.4719,
      "eval_samples_per_second": 1077.841,
      "eval_steps_per_second": 4.241,
      "step": 2970
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.7074270248413086,
      "learning_rate": 6.840085287846483e-05,
      "loss": 2.9631,
      "step": 3465
    },
    {
      "epoch": 7.0,
      "eval_loss": 3.1518795490264893,
      "eval_runtime": 30.1176,
      "eval_samples_per_second": 1054.732,
      "eval_steps_per_second": 4.15,
      "step": 3465
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.7585396766662598,
      "learning_rate": 6.312366737739872e-05,
      "loss": 2.901,
      "step": 3960
    },
    {
      "epoch": 8.0,
      "eval_loss": 3.137213706970215,
      "eval_runtime": 29.9134,
      "eval_samples_per_second": 1061.933,
      "eval_steps_per_second": 4.179,
      "step": 3960
    },
    {
      "epoch": 9.0,
      "grad_norm": 3.954860210418701,
      "learning_rate": 5.784648187633263e-05,
      "loss": 2.8443,
      "step": 4455
    },
    {
      "epoch": 9.0,
      "eval_loss": 3.1181836128234863,
      "eval_runtime": 29.5098,
      "eval_samples_per_second": 1076.455,
      "eval_steps_per_second": 4.236,
      "step": 4455
    },
    {
      "epoch": 10.0,
      "grad_norm": 4.154609680175781,
      "learning_rate": 5.256929637526653e-05,
      "loss": 2.7925,
      "step": 4950
    },
    {
      "epoch": 10.0,
      "eval_loss": 3.112929344177246,
      "eval_runtime": 30.5249,
      "eval_samples_per_second": 1040.659,
      "eval_steps_per_second": 4.095,
      "step": 4950
    },
    {
      "epoch": 11.0,
      "grad_norm": 4.463516712188721,
      "learning_rate": 4.7292110874200426e-05,
      "loss": 2.7436,
      "step": 5445
    },
    {
      "epoch": 11.0,
      "eval_loss": 3.110565185546875,
      "eval_runtime": 28.9255,
      "eval_samples_per_second": 1098.202,
      "eval_steps_per_second": 4.321,
      "step": 5445
    },
    {
      "epoch": 12.0,
      "grad_norm": 4.773754119873047,
      "learning_rate": 4.201492537313433e-05,
      "loss": 2.6967,
      "step": 5940
    },
    {
      "epoch": 12.0,
      "eval_loss": 3.1144516468048096,
      "eval_runtime": 30.5356,
      "eval_samples_per_second": 1040.296,
      "eval_steps_per_second": 4.094,
      "step": 5940
    },
    {
      "epoch": 13.0,
      "grad_norm": 5.04469108581543,
      "learning_rate": 3.673773987206823e-05,
      "loss": 2.6527,
      "step": 6435
    },
    {
      "epoch": 13.0,
      "eval_loss": 3.11479115486145,
      "eval_runtime": 29.5691,
      "eval_samples_per_second": 1074.296,
      "eval_steps_per_second": 4.227,
      "step": 6435
    },
    {
      "epoch": 14.0,
      "grad_norm": 5.253041744232178,
      "learning_rate": 3.1460554371002134e-05,
      "loss": 2.6122,
      "step": 6930
    },
    {
      "epoch": 14.0,
      "eval_loss": 3.120633840560913,
      "eval_runtime": 29.5965,
      "eval_samples_per_second": 1073.303,
      "eval_steps_per_second": 4.223,
      "step": 6930
    },
    {
      "epoch": 15.0,
      "grad_norm": 5.515427589416504,
      "learning_rate": 2.6183368869936037e-05,
      "loss": 2.575,
      "step": 7425
    },
    {
      "epoch": 15.0,
      "eval_loss": 3.1303443908691406,
      "eval_runtime": 29.5955,
      "eval_samples_per_second": 1073.34,
      "eval_steps_per_second": 4.224,
      "step": 7425
    },
    {
      "epoch": 16.0,
      "grad_norm": 6.251354694366455,
      "learning_rate": 2.0906183368869936e-05,
      "loss": 2.542,
      "step": 7920
    },
    {
      "epoch": 16.0,
      "eval_loss": 3.1400558948516846,
      "eval_runtime": 30.5232,
      "eval_samples_per_second": 1040.715,
      "eval_steps_per_second": 4.095,
      "step": 7920
    },
    {
      "epoch": 17.0,
      "grad_norm": Infinity,
      "learning_rate": 1.562899786780384e-05,
      "loss": 2.512,
      "step": 8415
    },
    {
      "epoch": 17.0,
      "eval_loss": 3.144320487976074,
      "eval_runtime": 29.8375,
      "eval_samples_per_second": 1064.633,
      "eval_steps_per_second": 4.189,
      "step": 8415
    }
  ],
  "logging_steps": 500,
  "max_steps": 9900,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "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": 1.41869684588544e+17,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}