trainer_state.json

{
  "best_metric": 0.9258666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0122/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": 4.683515548706055,
      "learning_rate": 4.8108254489432715e-05,
      "loss": 0.7463,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8792,
      "eval_loss": 0.39707738161087036,
      "eval_runtime": 6.3868,
      "eval_samples_per_second": 587.15,
      "eval_steps_per_second": 9.238,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.03202857822179794,
      "learning_rate": 4.2698504093359784e-05,
      "loss": 0.1943,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8978666666666667,
      "eval_loss": 0.3433379530906677,
      "eval_runtime": 6.1788,
      "eval_samples_per_second": 606.917,
      "eval_steps_per_second": 9.549,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.12075181305408478,
      "learning_rate": 3.459431688641196e-05,
      "loss": 0.0856,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9018666666666667,
      "eval_loss": 0.3407345414161682,
      "eval_runtime": 7.4396,
      "eval_samples_per_second": 504.059,
      "eval_steps_per_second": 7.931,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.02465425804257393,
      "learning_rate": 2.5029481906207065e-05,
      "loss": 0.0398,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9098666666666667,
      "eval_loss": 0.345005065202713,
      "eval_runtime": 6.3317,
      "eval_samples_per_second": 592.254,
      "eval_steps_per_second": 9.318,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.002609114395454526,
      "learning_rate": 1.5460158573036288e-05,
      "loss": 0.0174,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9202666666666667,
      "eval_loss": 0.32711267471313477,
      "eval_runtime": 6.3133,
      "eval_samples_per_second": 593.986,
      "eval_steps_per_second": 9.345,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0001535084447823465,
      "learning_rate": 7.343189618242854e-06,
      "loss": 0.0029,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9250666666666667,
      "eval_loss": 0.31554797291755676,
      "eval_runtime": 6.2346,
      "eval_samples_per_second": 601.482,
      "eval_steps_per_second": 9.463,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.10261183232069016,
      "learning_rate": 1.91430998468726e-06,
      "loss": 0.0004,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9258666666666666,
      "eval_loss": 0.31075194478034973,
      "eval_runtime": 6.3032,
      "eval_samples_per_second": 594.932,
      "eval_steps_per_second": 9.36,
      "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
}