trainer_state.json

{
  "best_metric": 0.8776,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0997/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.4394048750400543,
      "learning_rate": 0.0003,
      "loss": 0.5783,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.5045955777168274,
      "eval_runtime": 5.9353,
      "eval_samples_per_second": 631.808,
      "eval_steps_per_second": 9.94,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 0.2732,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8752,
      "eval_loss": 0.46462154388427734,
      "eval_runtime": 5.9304,
      "eval_samples_per_second": 632.34,
      "eval_steps_per_second": 9.949,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.017603827640414238,
      "learning_rate": 0.0003,
      "loss": 0.1935,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8776,
      "eval_loss": 0.4568835198879242,
      "eval_runtime": 5.8785,
      "eval_samples_per_second": 637.915,
      "eval_steps_per_second": 10.037,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 1332,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 4.94223961867776e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}