smids_10x_deit_small_sgd_001_fold4
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: 0.3291
- 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: 0.001
- 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 |
|---|---|---|---|---|
| 0.5469 | 1.0 | 750 | 0.5533 | 0.7983 |
| 0.4148 | 2.0 | 1500 | 0.4326 | 0.8367 |
| 0.3982 | 3.0 | 2250 | 0.3912 | 0.8467 |
| 0.355 | 4.0 | 3000 | 0.3693 | 0.8533 |
| 0.3032 | 5.0 | 3750 | 0.3569 | 0.8583 |
| 0.2345 | 6.0 | 4500 | 0.3466 | 0.8617 |
| 0.2053 | 7.0 | 5250 | 0.3412 | 0.865 |
| 0.2443 | 8.0 | 6000 | 0.3381 | 0.8633 |
| 0.259 | 9.0 | 6750 | 0.3314 | 0.875 |
| 0.2146 | 10.0 | 7500 | 0.3275 | 0.8717 |
| 0.2301 | 11.0 | 8250 | 0.3262 | 0.8733 |
| 0.298 | 12.0 | 9000 | 0.3264 | 0.8733 |
| 0.2031 | 13.0 | 9750 | 0.3234 | 0.8783 |
| 0.1941 | 14.0 | 10500 | 0.3276 | 0.8783 |
| 0.1822 | 15.0 | 11250 | 0.3209 | 0.88 |
| 0.2209 | 16.0 | 12000 | 0.3226 | 0.8767 |
| 0.1294 | 17.0 | 12750 | 0.3179 | 0.8817 |
| 0.1726 | 18.0 | 13500 | 0.3224 | 0.88 |
| 0.2222 | 19.0 | 14250 | 0.3196 | 0.8833 |
| 0.1604 | 20.0 | 15000 | 0.3199 | 0.8817 |
| 0.1742 | 21.0 | 15750 | 0.3204 | 0.8783 |
| 0.1599 | 22.0 | 16500 | 0.3188 | 0.88 |
| 0.1753 | 23.0 | 17250 | 0.3189 | 0.8817 |
| 0.1975 | 24.0 | 18000 | 0.3189 | 0.8817 |
| 0.1797 | 25.0 | 18750 | 0.3190 | 0.8817 |
| 0.1646 | 26.0 | 19500 | 0.3244 | 0.8817 |
| 0.1585 | 27.0 | 20250 | 0.3244 | 0.885 |
| 0.1303 | 28.0 | 21000 | 0.3225 | 0.8817 |
| 0.1144 | 29.0 | 21750 | 0.3207 | 0.8817 |
| 0.1409 | 30.0 | 22500 | 0.3230 | 0.8817 |
| 0.1303 | 31.0 | 23250 | 0.3219 | 0.8833 |
| 0.1405 | 32.0 | 24000 | 0.3260 | 0.8817 |
| 0.1503 | 33.0 | 24750 | 0.3248 | 0.88 |
| 0.1402 | 34.0 | 25500 | 0.3257 | 0.8817 |
| 0.1266 | 35.0 | 26250 | 0.3227 | 0.88 |
| 0.1495 | 36.0 | 27000 | 0.3271 | 0.8817 |
| 0.1021 | 37.0 | 27750 | 0.3248 | 0.8833 |
| 0.1616 | 38.0 | 28500 | 0.3242 | 0.885 |
| 0.158 | 39.0 | 29250 | 0.3254 | 0.88 |
| 0.1668 | 40.0 | 30000 | 0.3256 | 0.8833 |
| 0.1276 | 41.0 | 30750 | 0.3297 | 0.88 |
| 0.1072 | 42.0 | 31500 | 0.3307 | 0.88 |
| 0.1457 | 43.0 | 32250 | 0.3289 | 0.8783 |
| 0.1691 | 44.0 | 33000 | 0.3278 | 0.8817 |
| 0.1442 | 45.0 | 33750 | 0.3288 | 0.88 |
| 0.1231 | 46.0 | 34500 | 0.3279 | 0.88 |
| 0.1011 | 47.0 | 35250 | 0.3276 | 0.8767 |
| 0.1059 | 48.0 | 36000 | 0.3287 | 0.8767 |
| 0.1263 | 49.0 | 36750 | 0.3292 | 0.8767 |
| 0.1053 | 50.0 | 37500 | 0.3291 | 0.8767 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2