trainer_state.json

{
  "best_metric": 0.9378666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0381/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": 0.6367744207382202,
      "learning_rate": 0.0001,
      "loss": 0.5991,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9354666666666667,
      "eval_loss": 0.23996350169181824,
      "eval_runtime": 7.3342,
      "eval_samples_per_second": 511.302,
      "eval_steps_per_second": 8.044,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.16909027099609375,
      "learning_rate": 0.0001,
      "loss": 0.135,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.936,
      "eval_loss": 0.2288711965084076,
      "eval_runtime": 7.4825,
      "eval_samples_per_second": 501.17,
      "eval_steps_per_second": 7.885,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.035732779651880264,
      "learning_rate": 0.0001,
      "loss": 0.0793,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9378666666666666,
      "eval_loss": 0.23559844493865967,
      "eval_runtime": 7.422,
      "eval_samples_per_second": 505.252,
      "eval_steps_per_second": 7.949,
      "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
}