trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 30.0,
  "global_step": 1740,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8.62,
      "learning_rate": 0.0003563218390804598,
      "loss": 6.498,
      "step": 500
    },
    {
      "epoch": 17.24,
      "learning_rate": 0.00021264367816091953,
      "loss": 6.3597,
      "step": 1000
    },
    {
      "epoch": 25.86,
      "learning_rate": 6.896551724137931e-05,
      "loss": 6.3417,
      "step": 1500
    },
    {
      "epoch": 30.0,
      "step": 1740,
      "total_flos": 6403012579676160.0,
      "train_loss": 6.3856693662446125,
      "train_runtime": 2805.3129,
      "train_samples_per_second": 2.449,
      "train_steps_per_second": 0.62
    }
  ],
  "max_steps": 1740,
  "num_train_epochs": 30,
  "total_flos": 6403012579676160.0,
  "trial_name": null,
  "trial_params": null
}