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": 0.8628876209259033,
      "learning_rate": 1e-05,
      "loss": 2.3603,
      "step": 1
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.829474687576294,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.2712,
      "step": 2
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8325985074043274,
      "learning_rate": 6e-06,
      "loss": 2.3117,
      "step": 3
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8130093812942505,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.2507,
      "step": 4
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.8112923502922058,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.2536,
      "step": 5
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.2576935291290283,
      "eval_runtime": 15.1321,
      "eval_samples_per_second": 0.132,
      "eval_steps_per_second": 0.066,
      "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": 2.9434195784761344e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}