trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "global_step": 18,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.17,
      "learning_rate": 0.0001875,
      "loss": 1.2704,
      "step": 3
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.1318,
      "step": 6
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00011250000000000001,
      "loss": 0.9672,
      "step": 9
    },
    {
      "epoch": 0.67,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.8685,
      "step": 12
    },
    {
      "epoch": 0.83,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.7864,
      "step": 15
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0,
      "loss": 0.7414,
      "step": 18
    }
  ],
  "max_steps": 18,
  "num_train_epochs": 1,
  "total_flos": 1431942885015552.0,
  "trial_name": null,
  "trial_params": null
}