trainer_state.json

{
  "best_global_step": null,
  "best_metric": Infinity,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 167,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.29985007496251875,
      "grad_norm": NaN,
      "learning_rate": 0.0002,
      "loss": 0.0,
      "step": 50
    },
    {
      "epoch": 0.5997001499250375,
      "grad_norm": NaN,
      "learning_rate": 0.0002,
      "loss": 0.0,
      "step": 100
    },
    {
      "epoch": 0.8995502248875562,
      "grad_norm": NaN,
      "learning_rate": 0.0002,
      "loss": 0.0,
      "step": 150
    },
    {
      "epoch": 1.0,
      "eval_bleu": 0.055009227631916784,
      "eval_cer": 7.540007176607338,
      "eval_f1": 0.2276746762612516,
      "eval_loss": NaN,
      "eval_meteor": 0.1912918468665843,
      "eval_runtime": 1220.9223,
      "eval_samples_per_second": 0.819,
      "eval_steps_per_second": 0.069,
      "eval_wer": 4.724329088179987,
      "step": 167
    }
  ],
  "logging_steps": 50,
  "max_steps": 498,
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2019927351296e+16,
  "train_batch_size": 12,
  "trial_name": null,
  "trial_params": null
}