trainer_state.json

{
  "best_metric": 0.9101333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0242/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": 8.77055835723877,
      "learning_rate": 0.00047524221697560476,
      "loss": 0.8148,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8074666666666667,
      "eval_loss": 0.6755308508872986,
      "eval_runtime": 5.6593,
      "eval_samples_per_second": 662.629,
      "eval_steps_per_second": 10.425,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.5361967086792,
      "learning_rate": 0.0004058724504646834,
      "loss": 0.455,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8261333333333334,
      "eval_loss": 0.6050750017166138,
      "eval_runtime": 4.9882,
      "eval_samples_per_second": 751.773,
      "eval_steps_per_second": 11.828,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6115542054176331,
      "learning_rate": 0.0003056302334890786,
      "loss": 0.2815,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8496,
      "eval_loss": 0.5481494665145874,
      "eval_runtime": 4.9255,
      "eval_samples_per_second": 761.345,
      "eval_steps_per_second": 11.978,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.03009689413011074,
      "learning_rate": 0.00019436976651092142,
      "loss": 0.156,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8714666666666666,
      "eval_loss": 0.4591886103153229,
      "eval_runtime": 4.4541,
      "eval_samples_per_second": 841.929,
      "eval_steps_per_second": 13.246,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.17713093757629395,
      "learning_rate": 9.412754953531663e-05,
      "loss": 0.06,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8952,
      "eval_loss": 0.3990393877029419,
      "eval_runtime": 5.8748,
      "eval_samples_per_second": 638.32,
      "eval_steps_per_second": 10.043,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.07303586602210999,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 0.0166,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9074666666666666,
      "eval_loss": 0.3637605607509613,
      "eval_runtime": 4.4171,
      "eval_samples_per_second": 848.968,
      "eval_steps_per_second": 13.357,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.2705190269589615e-10,
      "loss": 0.0065,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9101333333333333,
      "eval_loss": 0.3668527901172638,
      "eval_runtime": 6.626,
      "eval_samples_per_second": 565.951,
      "eval_steps_per_second": 8.904,
      "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
}