last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9305019305019306,
  "eval_steps": 1000,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.97,
      "learning_rate": 9.613899613899614e-05,
      "loss": 2.291,
      "step": 500
    },
    {
      "epoch": 1.93,
      "learning_rate": 9.227799227799229e-05,
      "loss": 2.2939,
      "step": 1000
    },
    {
      "epoch": 1.93,
      "eval_loss": 2.303375720977783,
      "eval_runtime": 57.8181,
      "eval_samples_per_second": 50.555,
      "eval_steps_per_second": 3.165,
      "step": 1000
    }
  ],
  "logging_steps": 500,
  "max_steps": 12950,
  "num_train_epochs": 25,
  "save_steps": 1000,
  "total_flos": 8.660925478487654e+16,
  "trial_name": null,
  "trial_params": null
}