trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 24,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.875,
      "grad_norm": 1.7773577098978421e-07,
      "learning_rate": 1.3285352466760777e-05,
      "loss": 0.0,
      "step": 7
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.970873786407767,
      "eval_f1": 0.9708202209574825,
      "eval_loss": 0.3374857008457184,
      "eval_runtime": 1.5469,
      "eval_samples_per_second": 66.584,
      "eval_steps_per_second": 1.293,
      "step": 8
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.6692462168066413e-07,
      "learning_rate": 7.814913215741634e-06,
      "loss": 0.0,
      "step": 14
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.970873786407767,
      "eval_f1": 0.9708202209574825,
      "eval_loss": 0.33735620975494385,
      "eval_runtime": 1.5393,
      "eval_samples_per_second": 66.912,
      "eval_steps_per_second": 1.299,
      "step": 16
    },
    {
      "epoch": 2.625,
      "grad_norm": 1.698020923868171e-07,
      "learning_rate": 2.34447396472249e-06,
      "loss": 0.0,
      "step": 21
    }
  ],
  "logging_steps": 7,
  "max_steps": 24,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 47491256236032.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}