{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 11.0,
  "eval_steps": 500,
  "global_step": 5071,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.08,
      "grad_norm": 11.749343872070312,
      "learning_rate": 1.8028002366397162e-05,
      "loss": 0.3118,
      "step": 500
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.06206320971250534,
      "learning_rate": 1.6056004732794323e-05,
      "loss": 0.0983,
      "step": 1000
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.01628081686794758,
      "learning_rate": 1.4084007099191482e-05,
      "loss": 0.0414,
      "step": 1500
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.004006213042885065,
      "learning_rate": 1.2112009465588644e-05,
      "loss": 0.0188,
      "step": 2000
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.004198592156171799,
      "learning_rate": 1.0140011831985803e-05,
      "loss": 0.0082,
      "step": 2500
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.003632474457845092,
      "learning_rate": 8.168014198382964e-06,
      "loss": 0.0055,
      "step": 3000
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.047922104597091675,
      "learning_rate": 6.196016564780123e-06,
      "loss": 0.0063,
      "step": 3500
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.000518262677360326,
      "learning_rate": 4.224018931177283e-06,
      "loss": 0.002,
      "step": 4000
    },
    {
      "epoch": 9.76,
      "grad_norm": 0.0003728394804056734,
      "learning_rate": 2.252021297574443e-06,
      "loss": 0.001,
      "step": 4500
    },
    {
      "epoch": 10.85,
      "grad_norm": 0.0012362179113551974,
      "learning_rate": 2.8002366397160327e-07,
      "loss": 0.0009,
      "step": 5000
    },
    {
      "epoch": 11.0,
      "step": 5071,
      "total_flos": 1.511853612099748e+17,
      "train_loss": 0.04872322637245083,
      "train_runtime": 10725.6356,
      "train_samples_per_second": 15.125,
      "train_steps_per_second": 0.473
    }
  ],
  "logging_steps": 500,
  "max_steps": 5071,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 11,
  "save_steps": 500,
  "total_flos": 1.511853612099748e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}