trainer_state.json

{
  "best_metric": 0.9008,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0245/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": Infinity,
      "learning_rate": 8.559608050575152e-05,
      "loss": 1.5507,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7845333333333333,
      "eval_loss": 0.7549332976341248,
      "eval_runtime": 7.504,
      "eval_samples_per_second": 499.737,
      "eval_steps_per_second": 7.863,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.426042556762695,
      "learning_rate": 7.31517728663805e-05,
      "loss": 0.5422,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8402666666666667,
      "eval_loss": 0.5352168083190918,
      "eval_runtime": 7.4757,
      "eval_samples_per_second": 501.622,
      "eval_steps_per_second": 7.892,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 5.5190743462682014e-05,
      "loss": 0.3332,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8578666666666667,
      "eval_loss": 0.4729527235031128,
      "eval_runtime": 7.4184,
      "eval_samples_per_second": 505.497,
      "eval_steps_per_second": 7.953,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 18.7794132232666,
      "learning_rate": 3.516402310353216e-05,
      "loss": 0.1955,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8728,
      "eval_loss": 0.415597140789032,
      "eval_runtime": 7.3929,
      "eval_samples_per_second": 507.242,
      "eval_steps_per_second": 7.981,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.10795783996582,
      "learning_rate": 1.708543859906803e-05,
      "loss": 0.1126,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.37986376881599426,
      "eval_runtime": 7.4525,
      "eval_samples_per_second": 503.188,
      "eval_steps_per_second": 7.917,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.4114382266998291,
      "learning_rate": 4.53567532968734e-06,
      "loss": 0.0639,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8968,
      "eval_loss": 0.35474589467048645,
      "eval_runtime": 7.4633,
      "eval_samples_per_second": 502.46,
      "eval_steps_per_second": 7.905,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 27.49689292907715,
      "learning_rate": 3.6782363693366185e-10,
      "loss": 0.0429,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9008,
      "eval_loss": 0.3567441403865814,
      "eval_runtime": 8.0251,
      "eval_samples_per_second": 467.283,
      "eval_steps_per_second": 7.352,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}