trainer_state.json

{
  "best_metric": 0.94,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0809/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.7454113960266113,
      "learning_rate": 3e-05,
      "loss": 1.1344,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9245333333333333,
      "eval_loss": 0.29382824897766113,
      "eval_runtime": 11.6444,
      "eval_samples_per_second": 322.044,
      "eval_steps_per_second": 5.067,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2284272164106369,
      "learning_rate": 3e-05,
      "loss": 0.1777,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9352,
      "eval_loss": 0.22039498388767242,
      "eval_runtime": 12.708,
      "eval_samples_per_second": 295.091,
      "eval_steps_per_second": 4.643,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9719551801681519,
      "learning_rate": 3e-05,
      "loss": 0.0726,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9384,
      "eval_loss": 0.21802596747875214,
      "eval_runtime": 12.4442,
      "eval_samples_per_second": 301.344,
      "eval_steps_per_second": 4.741,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.06803560256958,
      "learning_rate": 3e-05,
      "loss": 0.0288,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9370666666666667,
      "eval_loss": 0.24558617174625397,
      "eval_runtime": 11.884,
      "eval_samples_per_second": 315.552,
      "eval_steps_per_second": 4.965,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0748777985572815,
      "learning_rate": 3e-05,
      "loss": 0.0131,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.94,
      "eval_loss": 0.24672503769397736,
      "eval_runtime": 12.7588,
      "eval_samples_per_second": 293.915,
      "eval_steps_per_second": 4.624,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}