trainer_state.json

{
  "best_metric": 0.8714666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0071/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "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": 7e-05,
      "loss": 0.9085,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8621333333333333,
      "eval_loss": 0.4656936824321747,
      "eval_runtime": 6.0935,
      "eval_samples_per_second": 615.413,
      "eval_steps_per_second": 9.682,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.014366960152983665,
      "learning_rate": 7e-05,
      "loss": 0.3109,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8602666666666666,
      "eval_loss": 0.48297613859176636,
      "eval_runtime": 5.8377,
      "eval_samples_per_second": 642.378,
      "eval_steps_per_second": 10.107,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.471111923456192,
      "learning_rate": 7e-05,
      "loss": 0.2018,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8544,
      "eval_loss": 0.5437843799591064,
      "eval_runtime": 5.8723,
      "eval_samples_per_second": 638.59,
      "eval_steps_per_second": 10.047,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1936206817626953,
      "learning_rate": 7e-05,
      "loss": 0.1434,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.86,
      "eval_loss": 0.546466588973999,
      "eval_runtime": 6.0658,
      "eval_samples_per_second": 618.22,
      "eval_steps_per_second": 9.727,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.477880954742432,
      "learning_rate": 7e-05,
      "loss": 0.1171,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8696,
      "eval_loss": 0.5393248200416565,
      "eval_runtime": 5.8321,
      "eval_samples_per_second": 642.99,
      "eval_steps_per_second": 10.116,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 52.42342758178711,
      "learning_rate": 7e-05,
      "loss": 0.1025,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8714666666666666,
      "eval_loss": 0.534389317035675,
      "eval_runtime": 5.8688,
      "eval_samples_per_second": 638.97,
      "eval_steps_per_second": 10.053,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}