smids_1x_deit_tiny_adamax_0001_fold4
This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.0781
- Accuracy: 0.8733
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 |
|---|---|---|---|---|
| 0.437 | 1.0 | 75 | 0.3878 | 0.8467 |
| 0.3579 | 2.0 | 150 | 0.3600 | 0.8533 |
| 0.205 | 3.0 | 225 | 0.4349 | 0.8517 |
| 0.1111 | 4.0 | 300 | 0.4456 | 0.8633 |
| 0.0776 | 5.0 | 375 | 0.5350 | 0.8483 |
| 0.0492 | 6.0 | 450 | 0.6337 | 0.845 |
| 0.046 | 7.0 | 525 | 0.6226 | 0.8567 |
| 0.0111 | 8.0 | 600 | 0.7050 | 0.8617 |
| 0.0753 | 9.0 | 675 | 0.7254 | 0.8767 |
| 0.0006 | 10.0 | 750 | 0.8282 | 0.8717 |
| 0.0291 | 11.0 | 825 | 0.8528 | 0.87 |
| 0.0019 | 12.0 | 900 | 0.8610 | 0.875 |
| 0.0132 | 13.0 | 975 | 0.9173 | 0.87 |
| 0.0003 | 14.0 | 1050 | 0.9559 | 0.8717 |
| 0.0002 | 15.0 | 1125 | 0.9248 | 0.875 |
| 0.0004 | 16.0 | 1200 | 1.0147 | 0.8717 |
| 0.0016 | 17.0 | 1275 | 0.9350 | 0.8733 |
| 0.0151 | 18.0 | 1350 | 1.0934 | 0.8667 |
| 0.0007 | 19.0 | 1425 | 0.9772 | 0.8733 |
| 0.0001 | 20.0 | 1500 | 0.9904 | 0.8683 |
| 0.0007 | 21.0 | 1575 | 1.0162 | 0.8783 |
| 0.0 | 22.0 | 1650 | 1.0249 | 0.8767 |
| 0.0 | 23.0 | 1725 | 1.0244 | 0.8767 |
| 0.0092 | 24.0 | 1800 | 1.0296 | 0.88 |
| 0.0033 | 25.0 | 1875 | 1.0414 | 0.8767 |
| 0.0 | 26.0 | 1950 | 1.0492 | 0.8783 |
| 0.0 | 27.0 | 2025 | 1.0373 | 0.8733 |
| 0.0041 | 28.0 | 2100 | 1.0586 | 0.8767 |
| 0.0098 | 29.0 | 2175 | 1.0478 | 0.8783 |
| 0.0 | 30.0 | 2250 | 1.0448 | 0.8767 |
| 0.0 | 31.0 | 2325 | 1.0423 | 0.8733 |
| 0.0 | 32.0 | 2400 | 1.0519 | 0.875 |
| 0.0 | 33.0 | 2475 | 1.0303 | 0.875 |
| 0.0 | 34.0 | 2550 | 1.0338 | 0.875 |
| 0.0 | 35.0 | 2625 | 1.0649 | 0.8767 |
| 0.0 | 36.0 | 2700 | 1.0482 | 0.8767 |
| 0.0047 | 37.0 | 2775 | 1.0661 | 0.875 |
| 0.0 | 38.0 | 2850 | 1.0529 | 0.875 |
| 0.0 | 39.0 | 2925 | 1.0563 | 0.875 |
| 0.0 | 40.0 | 3000 | 1.0657 | 0.875 |
| 0.0035 | 41.0 | 3075 | 1.0671 | 0.875 |
| 0.0 | 42.0 | 3150 | 1.0658 | 0.875 |
| 0.0 | 43.0 | 3225 | 1.0667 | 0.8733 |
| 0.0 | 44.0 | 3300 | 1.0735 | 0.875 |
| 0.0 | 45.0 | 3375 | 1.0750 | 0.875 |
| 0.0 | 46.0 | 3450 | 1.0752 | 0.8733 |
| 0.0 | 47.0 | 3525 | 1.0774 | 0.8733 |
| 0.0 | 48.0 | 3600 | 1.0784 | 0.8733 |
| 0.0 | 49.0 | 3675 | 1.0781 | 0.8733 |
| 0.0 | 50.0 | 3750 | 1.0781 | 0.8733 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0