trainer_state.json

{
  "best_metric": 1.685868740081787,
  "best_model_checkpoint": "saves/llama2_7b_test_4/checkpoint-8",
  "epoch": 0.08888888888888889,
  "eval_steps": 4,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008888888888888889,
      "grad_norm": 0.9656311869621277,
      "learning_rate": 0.0001,
      "loss": 1.469,
      "step": 1
    },
    {
      "epoch": 0.017777777777777778,
      "grad_norm": 2.022883653640747,
      "learning_rate": 9.698463103929542e-05,
      "loss": 2.9751,
      "step": 2
    },
    {
      "epoch": 0.02666666666666667,
      "grad_norm": 1.2307357788085938,
      "learning_rate": 8.83022221559489e-05,
      "loss": 1.8521,
      "step": 3
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 2.503842830657959,
      "learning_rate": 7.500000000000001e-05,
      "loss": 2.5627,
      "step": 4
    },
    {
      "epoch": 0.035555555555555556,
      "eval_accuracy": 0.6939260670748238,
      "eval_loss": 1.9984581470489502,
      "eval_runtime": 20.9951,
      "eval_samples_per_second": 4.763,
      "eval_steps_per_second": 4.763,
      "step": 4
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 1.6151976585388184,
      "learning_rate": 5.868240888334653e-05,
      "loss": 1.7733,
      "step": 5
    },
    {
      "epoch": 0.05333333333333334,
      "grad_norm": 1.3629400730133057,
      "learning_rate": 4.131759111665349e-05,
      "loss": 1.5353,
      "step": 6
    },
    {
      "epoch": 0.06222222222222222,
      "grad_norm": 2.5036630630493164,
      "learning_rate": 2.500000000000001e-05,
      "loss": 1.3856,
      "step": 7
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 2.502861499786377,
      "learning_rate": 1.1697777844051105e-05,
      "loss": 2.2846,
      "step": 8
    },
    {
      "epoch": 0.07111111111111111,
      "eval_accuracy": 0.6973038730142216,
      "eval_loss": 1.685868740081787,
      "eval_runtime": 21.1016,
      "eval_samples_per_second": 4.739,
      "eval_steps_per_second": 4.739,
      "step": 8
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.4686763286590576,
      "learning_rate": 3.0153689607045845e-06,
      "loss": 1.6419,
      "step": 9
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 1.3143010139465332,
      "learning_rate": 0.0,
      "loss": 1.367,
      "step": 10
    },
    {
      "epoch": 0.08888888888888889,
      "step": 10,
      "total_flos": 591688209530880.0,
      "train_loss": 1.8846651792526246,
      "train_runtime": 97.4531,
      "train_samples_per_second": 0.821,
      "train_steps_per_second": 0.103
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 4,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 591688209530880.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}