trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.6666666666666665,
  "eval_steps": 25,
  "global_step": 4,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 25.628726959228516,
      "learning_rate": 0.00015000000000000001,
      "loss": 70.8207,
      "step": 1
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 37.944496154785156,
      "learning_rate": 0.0001,
      "loss": 66.5958,
      "step": 2
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 24.337867736816406,
      "learning_rate": 5e-05,
      "loss": 69.4538,
      "step": 3
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 32.62247085571289,
      "learning_rate": 0.0,
      "loss": 61.7436,
      "step": 4
    },
    {
      "epoch": 1.6666666666666665,
      "step": 4,
      "total_flos": 15373112741376.0,
      "train_loss": 67.15349578857422,
      "train_runtime": 193.0014,
      "train_samples_per_second": 0.435,
      "train_steps_per_second": 0.021
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 4,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 15373112741376.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}