trainer_state.json

{
  "best_metric": 0.9533333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0000/checkpoints/checkpoint-1332",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1332,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 13.635677337646484,
      "learning_rate": 4e-05,
      "loss": 0.8001,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9384,
      "eval_loss": 0.24901273846626282,
      "eval_runtime": 12.884,
      "eval_samples_per_second": 291.058,
      "eval_steps_per_second": 4.579,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3.0030030030030033e-05,
      "loss": 0.1075,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9448,
      "eval_loss": 0.20298892259597778,
      "eval_runtime": 11.5318,
      "eval_samples_per_second": 325.188,
      "eval_steps_per_second": 5.116,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5611491799354553,
      "learning_rate": 2.003003003003003e-05,
      "loss": 0.0359,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9490666666666666,
      "eval_loss": 0.19206155836582184,
      "eval_runtime": 11.2318,
      "eval_samples_per_second": 333.873,
      "eval_steps_per_second": 5.253,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.06199859827756882,
      "learning_rate": 1.003003003003003e-05,
      "loss": 0.012,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9533333333333334,
      "eval_loss": 0.1961139589548111,
      "eval_runtime": 11.3355,
      "eval_samples_per_second": 330.818,
      "eval_steps_per_second": 5.205,
      "step": 1332
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.58965282490368e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}