trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.992,
  "eval_steps": 500,
  "global_step": 31,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.32,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 5.7397,
      "step": 10
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.2222222222222222e-05,
      "loss": 4.7137,
      "step": 20
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.111111111111111e-06,
      "loss": 4.2477,
      "step": 30
    },
    {
      "epoch": 0.99,
      "step": 31,
      "total_flos": 777619148046336.0,
      "train_loss": 4.877151981476815,
      "train_runtime": 551.4767,
      "train_samples_per_second": 7.253,
      "train_steps_per_second": 0.056
    }
  ],
  "logging_steps": 10,
  "max_steps": 31,
  "num_train_epochs": 1,
  "save_steps": 200.0,
  "total_flos": 777619148046336.0,
  "trial_name": null,
  "trial_params": null
}