trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 25,
  "global_step": 4,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.1477793455123901,
      "learning_rate": 0.00015000000000000001,
      "loss": 4.6942,
      "step": 1
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1735906600952148,
      "learning_rate": 0.0001,
      "loss": 4.6097,
      "step": 2
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.1909127235412598,
      "learning_rate": 5e-05,
      "loss": 4.5129,
      "step": 3
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.2604788541793823,
      "learning_rate": 0.0,
      "loss": 4.4442,
      "step": 4
    },
    {
      "epoch": 4.0,
      "step": 4,
      "total_flos": 1773820700928.0,
      "train_loss": 4.5652196407318115,
      "train_runtime": 30.0618,
      "train_samples_per_second": 1.064,
      "train_steps_per_second": 0.133
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 4,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 1773820700928.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}