trainer_state.json

{
  "best_metric": 0.9416,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0908/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": 0.6856261491775513,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.6244,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9213333333333333,
      "eval_loss": 0.2605384886264801,
      "eval_runtime": 6.2161,
      "eval_samples_per_second": 603.272,
      "eval_steps_per_second": 9.491,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.08505792170763016,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.1397,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9184,
      "eval_loss": 0.2877260446548462,
      "eval_runtime": 6.5794,
      "eval_samples_per_second": 569.963,
      "eval_steps_per_second": 8.967,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.24413323402404785,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.0657,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9213333333333333,
      "eval_loss": 0.2997373342514038,
      "eval_runtime": 6.3401,
      "eval_samples_per_second": 591.471,
      "eval_steps_per_second": 9.306,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.018408725038170815,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.0282,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9344,
      "eval_loss": 0.2629348933696747,
      "eval_runtime": 6.2843,
      "eval_samples_per_second": 596.722,
      "eval_steps_per_second": 9.388,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.510996341705322,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.0098,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.94,
      "eval_loss": 0.26203635334968567,
      "eval_runtime": 6.0998,
      "eval_samples_per_second": 614.775,
      "eval_steps_per_second": 9.672,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.009806605987250805,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.0023,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9402666666666667,
      "eval_loss": 0.27083009481430054,
      "eval_runtime": 6.082,
      "eval_samples_per_second": 616.573,
      "eval_steps_per_second": 9.701,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.1731488851999511e-05,
      "loss": 0.0012,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9416,
      "eval_loss": 0.26831552386283875,
      "eval_runtime": 6.2533,
      "eval_samples_per_second": 599.68,
      "eval_steps_per_second": 9.435,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}