trainer_state.json

{
  "best_metric": 0.9269333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0987/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "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": 4.447781114447781e-05,
      "loss": 0.837,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8648,
      "eval_loss": 0.4471435248851776,
      "eval_runtime": 13.1033,
      "eval_samples_per_second": 286.187,
      "eval_steps_per_second": 4.503,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 223.89096069335938,
      "learning_rate": 3.892225558892226e-05,
      "loss": 0.2249,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8834666666666666,
      "eval_loss": 0.3977927267551422,
      "eval_runtime": 13.0512,
      "eval_samples_per_second": 287.331,
      "eval_steps_per_second": 4.521,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.3943936824798584,
      "learning_rate": 3.33667000333667e-05,
      "loss": 0.1076,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8922666666666667,
      "eval_loss": 0.3943991959095001,
      "eval_runtime": 12.9213,
      "eval_samples_per_second": 290.219,
      "eval_steps_per_second": 4.566,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.21548227965831757,
      "learning_rate": 2.7811144477811146e-05,
      "loss": 0.0672,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.896,
      "eval_loss": 0.3870258033275604,
      "eval_runtime": 12.5327,
      "eval_samples_per_second": 299.218,
      "eval_steps_per_second": 4.708,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.4656703472137451,
      "learning_rate": 2.225558892225559e-05,
      "loss": 0.0343,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9101333333333333,
      "eval_loss": 0.34990161657333374,
      "eval_runtime": 12.8812,
      "eval_samples_per_second": 291.123,
      "eval_steps_per_second": 4.58,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.11994336545467377,
      "learning_rate": 1.6700033366700034e-05,
      "loss": 0.015,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9168,
      "eval_loss": 0.35074537992477417,
      "eval_runtime": 12.1522,
      "eval_samples_per_second": 308.586,
      "eval_steps_per_second": 4.855,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1286989152431488,
      "learning_rate": 1.1144477811144478e-05,
      "loss": 0.0072,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.924,
      "eval_loss": 0.3319821059703827,
      "eval_runtime": 13.0817,
      "eval_samples_per_second": 286.659,
      "eval_steps_per_second": 4.51,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.0018328744918107986,
      "learning_rate": 5.5889222555889224e-06,
      "loss": 0.003,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9269333333333334,
      "eval_loss": 0.33426710963249207,
      "eval_runtime": 13.0315,
      "eval_samples_per_second": 287.765,
      "eval_steps_per_second": 4.527,
      "step": 2664
    }
  ],
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}