trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 5,
  "global_step": 5,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "grad_norm": 1.3018276691436768,
      "learning_rate": 1e-05,
      "loss": 2.6559,
      "step": 1
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.1980626583099365,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.5666,
      "step": 2
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.2069164514541626,
      "learning_rate": 6e-06,
      "loss": 2.6111,
      "step": 3
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.1335670948028564,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.5498,
      "step": 4
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.1290104389190674,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.5509,
      "step": 5
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.558208465576172,
      "eval_runtime": 1.1848,
      "eval_samples_per_second": 1.688,
      "eval_steps_per_second": 0.844,
      "step": 5
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 5,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3912368186720256.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}