trainer_state.json

{
  "best_metric": 0.924,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0613/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": 0.5667551755905151,
      "learning_rate": 4.850126564054637e-05,
      "loss": 0.7541,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8701333333333333,
      "eval_loss": 0.39831969141960144,
      "eval_runtime": 11.8354,
      "eval_samples_per_second": 316.847,
      "eval_steps_per_second": 4.985,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.12144660949707,
      "learning_rate": 4.416794553834106e-05,
      "loss": 0.207,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8872,
      "eval_loss": 0.3826764225959778,
      "eval_runtime": 11.8722,
      "eval_samples_per_second": 315.864,
      "eval_steps_per_second": 4.97,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 7.492955684661865,
      "learning_rate": 3.7522688315464166e-05,
      "loss": 0.1047,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8970666666666667,
      "eval_loss": 0.37864211201667786,
      "eval_runtime": 11.5708,
      "eval_samples_per_second": 324.092,
      "eval_steps_per_second": 5.099,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.3499596118927,
      "learning_rate": 2.936701006654613e-05,
      "loss": 0.0495,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9090666666666667,
      "eval_loss": 0.33882755041122437,
      "eval_runtime": 11.4587,
      "eval_samples_per_second": 327.261,
      "eval_steps_per_second": 5.149,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.014926210977137089,
      "learning_rate": 2.0684605953398183e-05,
      "loss": 0.0182,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9090666666666667,
      "eval_loss": 0.3939250409603119,
      "eval_runtime": 11.4796,
      "eval_samples_per_second": 326.666,
      "eval_steps_per_second": 5.14,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0003937547735404223,
      "learning_rate": 1.2522702050703538e-05,
      "loss": 0.0068,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9194666666666667,
      "eval_loss": 0.34596219658851624,
      "eval_runtime": 11.5067,
      "eval_samples_per_second": 325.898,
      "eval_steps_per_second": 5.127,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.000868489732965827,
      "learning_rate": 5.865744425999756e-06,
      "loss": 0.001,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.924,
      "eval_loss": 0.3319045305252075,
      "eval_runtime": 13.7808,
      "eval_samples_per_second": 272.117,
      "eval_steps_per_second": 4.281,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}