trainer_state.json

{
  "best_metric": 0.3738666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0682/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": 26.83074188232422,
      "learning_rate": 0.0003,
      "loss": 3.4065,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.1632,
      "eval_loss": 3.1298155784606934,
      "eval_runtime": 10.6318,
      "eval_samples_per_second": 352.715,
      "eval_steps_per_second": 5.549,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 21.332796096801758,
      "learning_rate": 0.0003,
      "loss": 2.8803,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.24133333333333334,
      "eval_loss": 2.8620989322662354,
      "eval_runtime": 10.6811,
      "eval_samples_per_second": 351.087,
      "eval_steps_per_second": 5.524,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 27.66588020324707,
      "learning_rate": 0.0003,
      "loss": 2.6947,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.2869333333333333,
      "eval_loss": 2.6816635131835938,
      "eval_runtime": 10.2908,
      "eval_samples_per_second": 364.404,
      "eval_steps_per_second": 5.733,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 11.06808090209961,
      "learning_rate": 0.0003,
      "loss": 2.5159,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.30293333333333333,
      "eval_loss": 2.6478219032287598,
      "eval_runtime": 10.5448,
      "eval_samples_per_second": 355.626,
      "eval_steps_per_second": 5.595,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 15.589930534362793,
      "learning_rate": 0.0003,
      "loss": 2.3821,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.33386666666666664,
      "eval_loss": 2.4781205654144287,
      "eval_runtime": 10.5964,
      "eval_samples_per_second": 353.895,
      "eval_steps_per_second": 5.568,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 37.123321533203125,
      "learning_rate": 0.0003,
      "loss": 2.2203,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3488,
      "eval_loss": 2.409799575805664,
      "eval_runtime": 11.0164,
      "eval_samples_per_second": 340.403,
      "eval_steps_per_second": 5.356,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 2.0766,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3738666666666667,
      "eval_loss": 2.3308756351470947,
      "eval_runtime": 10.3387,
      "eval_samples_per_second": 362.714,
      "eval_steps_per_second": 5.707,
      "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
}