trainer_state.json

{
  "best_metric": 0.9384,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0454/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": 2.6272522522522523e-05,
      "loss": 0.8542,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8858666666666667,
      "eval_loss": 0.3666590750217438,
      "eval_runtime": 6.8641,
      "eval_samples_per_second": 546.319,
      "eval_steps_per_second": 8.595,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3129681646823883,
      "learning_rate": 2.2522522522522523e-05,
      "loss": 0.1909,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9093333333333333,
      "eval_loss": 0.30093076825141907,
      "eval_runtime": 6.9492,
      "eval_samples_per_second": 539.628,
      "eval_steps_per_second": 8.49,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 104.85924530029297,
      "learning_rate": 1.8772522522522523e-05,
      "loss": 0.0785,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9173333333333333,
      "eval_loss": 0.29062578082084656,
      "eval_runtime": 6.77,
      "eval_samples_per_second": 553.915,
      "eval_steps_per_second": 8.715,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.006828392390161753,
      "learning_rate": 1.5022522522522522e-05,
      "loss": 0.035,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9258666666666666,
      "eval_loss": 0.2750960886478424,
      "eval_runtime": 6.618,
      "eval_samples_per_second": 566.637,
      "eval_steps_per_second": 8.915,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0031280498951673508,
      "learning_rate": 1.1272522522522524e-05,
      "loss": 0.0147,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9296,
      "eval_loss": 0.2785312533378601,
      "eval_runtime": 6.9651,
      "eval_samples_per_second": 538.395,
      "eval_steps_per_second": 8.471,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0031017446890473366,
      "learning_rate": 7.522522522522523e-06,
      "loss": 0.0044,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.928,
      "eval_loss": 0.2919682562351227,
      "eval_runtime": 6.9071,
      "eval_samples_per_second": 542.917,
      "eval_steps_per_second": 8.542,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.013746746815741062,
      "learning_rate": 3.772522522522523e-06,
      "loss": 0.0021,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9384,
      "eval_loss": 0.26187729835510254,
      "eval_runtime": 6.8866,
      "eval_samples_per_second": 544.538,
      "eval_steps_per_second": 8.567,
      "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
}