smids_5x_deit_small_sgd_0001_fold1
This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.5286
- Accuracy: 0.7913
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.0382 | 1.0 | 376 | 1.0714 | 0.4090 |
| 1.0394 | 2.0 | 752 | 1.0356 | 0.4491 |
| 0.9541 | 3.0 | 1128 | 1.0000 | 0.5042 |
| 0.928 | 4.0 | 1504 | 0.9637 | 0.5559 |
| 0.8813 | 5.0 | 1880 | 0.9285 | 0.5943 |
| 0.8344 | 6.0 | 2256 | 0.8967 | 0.6227 |
| 0.7873 | 7.0 | 2632 | 0.8681 | 0.6361 |
| 0.8127 | 8.0 | 3008 | 0.8414 | 0.6544 |
| 0.8077 | 9.0 | 3384 | 0.8166 | 0.6628 |
| 0.7559 | 10.0 | 3760 | 0.7932 | 0.6811 |
| 0.7293 | 11.0 | 4136 | 0.7713 | 0.6962 |
| 0.7184 | 12.0 | 4512 | 0.7510 | 0.7028 |
| 0.6866 | 13.0 | 4888 | 0.7320 | 0.7078 |
| 0.6588 | 14.0 | 5264 | 0.7145 | 0.7195 |
| 0.6425 | 15.0 | 5640 | 0.6984 | 0.7245 |
| 0.6378 | 16.0 | 6016 | 0.6836 | 0.7312 |
| 0.5876 | 17.0 | 6392 | 0.6699 | 0.7396 |
| 0.6379 | 18.0 | 6768 | 0.6573 | 0.7429 |
| 0.6063 | 19.0 | 7144 | 0.6456 | 0.7479 |
| 0.5557 | 20.0 | 7520 | 0.6350 | 0.7496 |
| 0.5709 | 21.0 | 7896 | 0.6253 | 0.7513 |
| 0.5404 | 22.0 | 8272 | 0.6166 | 0.7563 |
| 0.5599 | 23.0 | 8648 | 0.6082 | 0.7529 |
| 0.5567 | 24.0 | 9024 | 0.6008 | 0.7613 |
| 0.5445 | 25.0 | 9400 | 0.5938 | 0.7646 |
| 0.5273 | 26.0 | 9776 | 0.5874 | 0.7629 |
| 0.5187 | 27.0 | 10152 | 0.5814 | 0.7613 |
| 0.4686 | 28.0 | 10528 | 0.5760 | 0.7629 |
| 0.502 | 29.0 | 10904 | 0.5710 | 0.7629 |
| 0.5086 | 30.0 | 11280 | 0.5663 | 0.7663 |
| 0.5383 | 31.0 | 11656 | 0.5621 | 0.7679 |
| 0.5306 | 32.0 | 12032 | 0.5581 | 0.7696 |
| 0.4719 | 33.0 | 12408 | 0.5545 | 0.7713 |
| 0.4733 | 34.0 | 12784 | 0.5512 | 0.7763 |
| 0.4916 | 35.0 | 13160 | 0.5482 | 0.7796 |
| 0.4659 | 36.0 | 13536 | 0.5454 | 0.7796 |
| 0.4447 | 37.0 | 13912 | 0.5429 | 0.7830 |
| 0.5196 | 38.0 | 14288 | 0.5406 | 0.7830 |
| 0.4685 | 39.0 | 14664 | 0.5386 | 0.7830 |
| 0.4526 | 40.0 | 15040 | 0.5367 | 0.7830 |
| 0.4896 | 41.0 | 15416 | 0.5350 | 0.7863 |
| 0.4446 | 42.0 | 15792 | 0.5336 | 0.7863 |
| 0.4328 | 43.0 | 16168 | 0.5323 | 0.7863 |
| 0.5156 | 44.0 | 16544 | 0.5312 | 0.7880 |
| 0.4252 | 45.0 | 16920 | 0.5303 | 0.7896 |
| 0.4576 | 46.0 | 17296 | 0.5296 | 0.7896 |
| 0.4261 | 47.0 | 17672 | 0.5291 | 0.7913 |
| 0.4841 | 48.0 | 18048 | 0.5288 | 0.7913 |
| 0.4563 | 49.0 | 18424 | 0.5286 | 0.7913 |
| 0.4361 | 50.0 | 18800 | 0.5286 | 0.7913 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2