trainer_state.json

{
  "best_metric": 0.30533333333333335,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0879/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": 0.0005,
      "loss": 3.48,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.18373333333333333,
      "eval_loss": 3.0908360481262207,
      "eval_runtime": 14.9337,
      "eval_samples_per_second": 251.111,
      "eval_steps_per_second": 3.951,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 23.010967254638672,
      "learning_rate": 0.0005,
      "loss": 3.0245,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2176,
      "eval_loss": 2.9312028884887695,
      "eval_runtime": 12.625,
      "eval_samples_per_second": 297.029,
      "eval_steps_per_second": 4.673,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 20.620025634765625,
      "learning_rate": 0.0005,
      "loss": 2.8622,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.23573333333333332,
      "eval_loss": 2.890881299972534,
      "eval_runtime": 11.4531,
      "eval_samples_per_second": 327.422,
      "eval_steps_per_second": 5.151,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 18.312299728393555,
      "learning_rate": 0.0005,
      "loss": 2.7397,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.24933333333333332,
      "eval_loss": 2.8369030952453613,
      "eval_runtime": 10.8878,
      "eval_samples_per_second": 344.422,
      "eval_steps_per_second": 5.419,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 15.705731391906738,
      "learning_rate": 0.0005,
      "loss": 2.6623,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.2856,
      "eval_loss": 2.644413948059082,
      "eval_runtime": 10.7827,
      "eval_samples_per_second": 347.778,
      "eval_steps_per_second": 5.472,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 18.099206924438477,
      "learning_rate": 0.0005,
      "loss": 2.5505,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.2968,
      "eval_loss": 2.5969839096069336,
      "eval_runtime": 10.7428,
      "eval_samples_per_second": 349.072,
      "eval_steps_per_second": 5.492,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 14.623054504394531,
      "learning_rate": 0.0005,
      "loss": 2.4716,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.30533333333333335,
      "eval_loss": 2.597694158554077,
      "eval_runtime": 10.9313,
      "eval_samples_per_second": 343.053,
      "eval_steps_per_second": 5.397,
      "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
}