smids_1x_deit_tiny_adamax_0001_fold5
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: 0.8318
- Accuracy: 0.8767
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 |
|---|---|---|---|---|
| 0.4622 | 1.0 | 75 | 0.3902 | 0.85 |
| 0.3997 | 2.0 | 150 | 0.3259 | 0.8767 |
| 0.1811 | 3.0 | 225 | 0.3681 | 0.8767 |
| 0.2239 | 4.0 | 300 | 0.3692 | 0.8717 |
| 0.1061 | 5.0 | 375 | 0.3361 | 0.89 |
| 0.0643 | 6.0 | 450 | 0.4246 | 0.8867 |
| 0.0349 | 7.0 | 525 | 0.5503 | 0.865 |
| 0.0577 | 8.0 | 600 | 0.4923 | 0.89 |
| 0.0047 | 9.0 | 675 | 0.6463 | 0.875 |
| 0.0074 | 10.0 | 750 | 0.6530 | 0.8683 |
| 0.037 | 11.0 | 825 | 0.6041 | 0.8767 |
| 0.0089 | 12.0 | 900 | 0.6572 | 0.8833 |
| 0.0191 | 13.0 | 975 | 0.7128 | 0.8867 |
| 0.0365 | 14.0 | 1050 | 0.8168 | 0.8733 |
| 0.0272 | 15.0 | 1125 | 0.7575 | 0.875 |
| 0.0001 | 16.0 | 1200 | 0.7279 | 0.8817 |
| 0.0112 | 17.0 | 1275 | 0.7893 | 0.885 |
| 0.0208 | 18.0 | 1350 | 0.8683 | 0.8667 |
| 0.0094 | 19.0 | 1425 | 0.8173 | 0.875 |
| 0.0001 | 20.0 | 1500 | 0.8675 | 0.8633 |
| 0.005 | 21.0 | 1575 | 0.8404 | 0.8633 |
| 0.0089 | 22.0 | 1650 | 0.8394 | 0.8633 |
| 0.0106 | 23.0 | 1725 | 0.7646 | 0.8733 |
| 0.0 | 24.0 | 1800 | 0.7970 | 0.8667 |
| 0.0166 | 25.0 | 1875 | 0.7834 | 0.8767 |
| 0.0 | 26.0 | 1950 | 0.7904 | 0.8833 |
| 0.0062 | 27.0 | 2025 | 0.8106 | 0.8667 |
| 0.0107 | 28.0 | 2100 | 0.7803 | 0.8767 |
| 0.0105 | 29.0 | 2175 | 0.7751 | 0.875 |
| 0.0047 | 30.0 | 2250 | 0.7593 | 0.875 |
| 0.0 | 31.0 | 2325 | 0.8560 | 0.8733 |
| 0.0 | 32.0 | 2400 | 0.7908 | 0.87 |
| 0.0066 | 33.0 | 2475 | 0.8221 | 0.875 |
| 0.0 | 34.0 | 2550 | 0.8079 | 0.8733 |
| 0.0052 | 35.0 | 2625 | 0.8173 | 0.875 |
| 0.0 | 36.0 | 2700 | 0.8258 | 0.8733 |
| 0.0045 | 37.0 | 2775 | 0.8236 | 0.8733 |
| 0.0 | 38.0 | 2850 | 0.8199 | 0.875 |
| 0.0 | 39.0 | 2925 | 0.8205 | 0.8733 |
| 0.0 | 40.0 | 3000 | 0.8239 | 0.8733 |
| 0.0036 | 41.0 | 3075 | 0.8262 | 0.8767 |
| 0.0 | 42.0 | 3150 | 0.8270 | 0.875 |
| 0.0074 | 43.0 | 3225 | 0.8280 | 0.875 |
| 0.0 | 44.0 | 3300 | 0.8301 | 0.875 |
| 0.0026 | 45.0 | 3375 | 0.8317 | 0.875 |
| 0.0 | 46.0 | 3450 | 0.8313 | 0.8767 |
| 0.0057 | 47.0 | 3525 | 0.8305 | 0.875 |
| 0.0054 | 48.0 | 3600 | 0.8318 | 0.8767 |
| 0.0 | 49.0 | 3675 | 0.8319 | 0.8767 |
| 0.0042 | 50.0 | 3750 | 0.8318 | 0.8767 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0