trainer_state.json

{
  "best_metric": 0.9525333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0432/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.8046677112579346,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.8536,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9242666666666667,
      "eval_loss": 0.2762451767921448,
      "eval_runtime": 10.5307,
      "eval_samples_per_second": 356.103,
      "eval_steps_per_second": 5.603,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5805880427360535,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.1237,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9384,
      "eval_loss": 0.22124387323856354,
      "eval_runtime": 10.7749,
      "eval_samples_per_second": 348.03,
      "eval_steps_per_second": 5.476,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.1372554749250412,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0447,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9421333333333334,
      "eval_loss": 0.21247252821922302,
      "eval_runtime": 10.6587,
      "eval_samples_per_second": 351.824,
      "eval_steps_per_second": 5.535,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.918959617614746,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.017,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9437333333333333,
      "eval_loss": 0.220442995429039,
      "eval_runtime": 10.5201,
      "eval_samples_per_second": 356.46,
      "eval_steps_per_second": 5.608,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.1201029047369957,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.0065,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9490666666666666,
      "eval_loss": 0.20500802993774414,
      "eval_runtime": 10.166,
      "eval_samples_per_second": 368.878,
      "eval_steps_per_second": 5.804,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.04130680859088898,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0026,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9514666666666667,
      "eval_loss": 0.20796631276607513,
      "eval_runtime": 10.823,
      "eval_samples_per_second": 346.485,
      "eval_steps_per_second": 5.451,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.1127794461127795e-05,
      "loss": 0.0016,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.952,
      "eval_loss": 0.2113092988729477,
      "eval_runtime": 11.1023,
      "eval_samples_per_second": 337.769,
      "eval_steps_per_second": 5.314,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.011887535452842712,
      "learning_rate": 5.572238905572239e-06,
      "loss": 0.0012,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9525333333333333,
      "eval_loss": 0.21260999143123627,
      "eval_runtime": 11.5317,
      "eval_samples_per_second": 325.192,
      "eval_steps_per_second": 5.116,
      "step": 2664
    }
  ],
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}