trainer_state.json

{
  "best_metric": 0.8888,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0539/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": 8.659485808097889e-05,
      "loss": 1.578,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7808,
      "eval_loss": 0.7690122127532959,
      "eval_runtime": 8.0206,
      "eval_samples_per_second": 467.544,
      "eval_steps_per_second": 7.356,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7027547359466553,
      "learning_rate": 7.685730736804761e-05,
      "loss": 0.5818,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8138666666666666,
      "eval_loss": 0.6134414076805115,
      "eval_runtime": 7.3397,
      "eval_samples_per_second": 510.922,
      "eval_steps_per_second": 8.039,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 6.231876231770601e-05,
      "loss": 0.3607,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8485333333333334,
      "eval_loss": 0.5359084010124207,
      "eval_runtime": 7.717,
      "eval_samples_per_second": 485.941,
      "eval_steps_per_second": 7.645,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.8492801189422607,
      "learning_rate": 4.510613478854492e-05,
      "loss": 0.2314,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8677333333333334,
      "eval_loss": 0.4495958089828491,
      "eval_runtime": 7.3959,
      "eval_samples_per_second": 507.039,
      "eval_steps_per_second": 7.977,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4252076745033264,
      "learning_rate": 2.7877349199942123e-05,
      "loss": 0.132,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8696,
      "eval_loss": 0.45762717723846436,
      "eval_runtime": 7.4915,
      "eval_samples_per_second": 500.568,
      "eval_steps_per_second": 7.876,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.042630743235349655,
      "learning_rate": 1.3255331978432106e-05,
      "loss": 0.0719,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8832,
      "eval_loss": 0.40288275480270386,
      "eval_runtime": 7.8509,
      "eval_samples_per_second": 477.655,
      "eval_steps_per_second": 7.515,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.05258096382021904,
      "learning_rate": 3.4661526970736e-06,
      "loss": 0.0374,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8888,
      "eval_loss": 0.40483880043029785,
      "eval_runtime": 7.4595,
      "eval_samples_per_second": 502.713,
      "eval_steps_per_second": 7.909,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}