trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1293,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.16,
      "learning_rate": 1.839907192575406e-05,
      "loss": 1.701,
      "step": 500
    },
    {
      "epoch": 2.32,
      "learning_rate": 6.798143851508122e-06,
      "loss": 1.1061,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "step": 1293,
      "total_flos": 3038946605028864.0,
      "train_loss": 1.263148379344306,
      "train_runtime": 9000.3857,
      "train_samples_per_second": 1.723,
      "train_steps_per_second": 0.144
    }
  ],
  "logging_steps": 500,
  "max_steps": 1293,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 3038946605028864.0,
  "trial_name": null,
  "trial_params": null
}