trainer_state.json

{
  "best_metric": 0.5941333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0230/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": 0.00047538820348083834,
      "loss": 3.2148,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.26293333333333335,
      "eval_loss": 2.738433837890625,
      "eval_runtime": 7.1122,
      "eval_samples_per_second": 527.26,
      "eval_steps_per_second": 8.296,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00040639873814655834,
      "loss": 2.4866,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3656,
      "eval_loss": 2.2688302993774414,
      "eval_runtime": 6.5525,
      "eval_samples_per_second": 572.298,
      "eval_steps_per_second": 9.004,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5495281219482422,
      "learning_rate": 0.00030628700751103485,
      "loss": 1.9279,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4632,
      "eval_loss": 1.904839277267456,
      "eval_runtime": 6.2869,
      "eval_samples_per_second": 596.479,
      "eval_steps_per_second": 9.385,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 26.40676498413086,
      "learning_rate": 0.00019502694472310563,
      "loss": 1.4044,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5354666666666666,
      "eval_loss": 1.6532715559005737,
      "eval_runtime": 6.3345,
      "eval_samples_per_second": 591.995,
      "eval_steps_per_second": 9.314,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 9.491910332815573e-05,
      "loss": 0.7864,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5746666666666667,
      "eval_loss": 1.5265676975250244,
      "eval_runtime": 6.2433,
      "eval_samples_per_second": 600.639,
      "eval_steps_per_second": 9.45,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.7270925045013428,
      "learning_rate": 2.519819627604078e-05,
      "loss": 0.27,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5882666666666667,
      "eval_loss": 1.5380734205245972,
      "eval_runtime": 6.7163,
      "eval_samples_per_second": 558.341,
      "eval_steps_per_second": 8.785,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.6822614669799805,
      "learning_rate": 2.0434646496314548e-09,
      "loss": 0.08,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5941333333333333,
      "eval_loss": 1.5440850257873535,
      "eval_runtime": 6.9275,
      "eval_samples_per_second": 541.322,
      "eval_steps_per_second": 8.517,
      "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
}