checkpoint-500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.384615384615385,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.08,
      "learning_rate": 0.0002,
      "loss": 2.728,
      "step": 100
    },
    {
      "epoch": 6.15,
      "learning_rate": 0.0002,
      "loss": 1.5741,
      "step": 200
    },
    {
      "epoch": 9.23,
      "learning_rate": 0.0002,
      "loss": 0.8154,
      "step": 300
    },
    {
      "epoch": 12.31,
      "learning_rate": 0.0002,
      "loss": 0.4314,
      "step": 400
    },
    {
      "epoch": 15.38,
      "learning_rate": 0.0002,
      "loss": 0.2123,
      "step": 500
    }
  ],
  "logging_steps": 100,
  "max_steps": 500,
  "num_train_epochs": 16,
  "save_steps": 100,
  "total_flos": 6.276242983098778e+16,
  "trial_name": null,
  "trial_params": null
}