trainer_state.json

{
  "best_metric": 0.8770666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0219/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 189.91331481933594,
      "learning_rate": 6.733578363789503e-05,
      "loss": 3.1802,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6869333333333333,
      "eval_loss": 1.5372601747512817,
      "eval_runtime": 5.228,
      "eval_samples_per_second": 717.296,
      "eval_steps_per_second": 11.285,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 170.17587280273438,
      "learning_rate": 5.974873734152915e-05,
      "loss": 0.936,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8138666666666666,
      "eval_loss": 0.6322967410087585,
      "eval_runtime": 5.4136,
      "eval_samples_per_second": 692.703,
      "eval_steps_per_second": 10.899,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 107.09070587158203,
      "learning_rate": 4.839392013277814e-05,
      "loss": 0.5367,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8533333333333334,
      "eval_loss": 0.47847461700439453,
      "eval_runtime": 5.1703,
      "eval_samples_per_second": 725.299,
      "eval_steps_per_second": 11.411,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 138.44371032714844,
      "learning_rate": 3.5e-05,
      "loss": 0.3977,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8666666666666667,
      "eval_loss": 0.44247642159461975,
      "eval_runtime": 5.143,
      "eval_samples_per_second": 729.14,
      "eval_steps_per_second": 11.472,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 84.33867645263672,
      "learning_rate": 2.1606079867221858e-05,
      "loss": 0.3266,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8677333333333334,
      "eval_loss": 0.42508846521377563,
      "eval_runtime": 5.1839,
      "eval_samples_per_second": 723.389,
      "eval_steps_per_second": 11.381,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 158.33152770996094,
      "learning_rate": 1.0251262658470837e-05,
      "loss": 0.2792,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8725333333333334,
      "eval_loss": 0.40501725673675537,
      "eval_runtime": 5.167,
      "eval_samples_per_second": 725.756,
      "eval_steps_per_second": 11.419,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 150.33273315429688,
      "learning_rate": 2.664216362104964e-06,
      "loss": 0.2591,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8770666666666667,
      "eval_loss": 0.4024164080619812,
      "eval_runtime": 5.1811,
      "eval_samples_per_second": 723.785,
      "eval_steps_per_second": 11.388,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}