trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "global_step": 4790,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.04,
      "learning_rate": 4.478079331941545e-05,
      "loss": 0.667,
      "step": 500
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.95615866388309e-05,
      "loss": 0.4544,
      "step": 1000
    },
    {
      "epoch": 3.13,
      "learning_rate": 3.434237995824635e-05,
      "loss": 0.4076,
      "step": 1500
    },
    {
      "epoch": 4.18,
      "learning_rate": 2.9123173277661797e-05,
      "loss": 0.3803,
      "step": 2000
    },
    {
      "epoch": 5.22,
      "learning_rate": 2.3903966597077245e-05,
      "loss": 0.3532,
      "step": 2500
    },
    {
      "epoch": 6.26,
      "learning_rate": 1.8684759916492694e-05,
      "loss": 0.338,
      "step": 3000
    },
    {
      "epoch": 7.31,
      "learning_rate": 1.3465553235908144e-05,
      "loss": 0.322,
      "step": 3500
    },
    {
      "epoch": 8.35,
      "learning_rate": 8.246346555323591e-06,
      "loss": 0.3139,
      "step": 4000
    },
    {
      "epoch": 9.39,
      "learning_rate": 3.02713987473904e-06,
      "loss": 0.305,
      "step": 4500
    },
    {
      "epoch": 10.0,
      "step": 4790,
      "total_flos": 6.57117365981184e+16,
      "train_loss": 0.3880401197206501,
      "train_runtime": 3696.5294,
      "train_samples_per_second": 12.958,
      "train_steps_per_second": 1.296
    }
  ],
  "max_steps": 4790,
  "num_train_epochs": 10,
  "total_flos": 6.57117365981184e+16,
  "trial_name": null,
  "trial_params": null
}