trainer_state.json

{
  "best_metric": 0.9152,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0850/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": 8.562221064561357e-05,
      "loss": 0.9258,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8349333333333333,
      "eval_loss": 0.5422833561897278,
      "eval_runtime": 13.3562,
      "eval_samples_per_second": 280.768,
      "eval_steps_per_second": 4.417,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 20.126869201660156,
      "learning_rate": 7.319906209778102e-05,
      "loss": 0.318,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8490666666666666,
      "eval_loss": 0.5308270454406738,
      "eval_runtime": 13.4974,
      "eval_samples_per_second": 277.831,
      "eval_steps_per_second": 4.371,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.572601556777954,
      "learning_rate": 5.5190743462682014e-05,
      "loss": 0.1566,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8589333333333333,
      "eval_loss": 0.5228258967399597,
      "eval_runtime": 12.7421,
      "eval_samples_per_second": 294.299,
      "eval_steps_per_second": 4.63,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1113591194152832,
      "learning_rate": 3.516402310353216e-05,
      "loss": 0.0727,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.884,
      "eval_loss": 0.43439069390296936,
      "eval_runtime": 13.1297,
      "eval_samples_per_second": 285.612,
      "eval_steps_per_second": 4.494,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7765991687774658,
      "learning_rate": 1.708543859906803e-05,
      "loss": 0.0239,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9114666666666666,
      "eval_loss": 0.375984787940979,
      "eval_runtime": 12.9767,
      "eval_samples_per_second": 288.98,
      "eval_steps_per_second": 4.547,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.6349732875823975,
      "learning_rate": 4.53567532968734e-06,
      "loss": 0.005,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9144,
      "eval_loss": 0.34683623909950256,
      "eval_runtime": 13.0015,
      "eval_samples_per_second": 288.428,
      "eval_steps_per_second": 4.538,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.03335009887814522,
      "learning_rate": 3.6782363693366185e-10,
      "loss": 0.0021,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9152,
      "eval_loss": 0.3568275570869446,
      "eval_runtime": 13.8873,
      "eval_samples_per_second": 270.032,
      "eval_steps_per_second": 4.248,
      "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
}