smids_3x_deit_tiny_sgd_00001_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: 1.1127
- Accuracy: 0.3917
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: 1e-05
- 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 |
|---|---|---|---|---|
| 1.3668 | 1.0 | 225 | 1.3432 | 0.3467 |
| 1.2646 | 2.0 | 450 | 1.3216 | 0.3467 |
| 1.3414 | 3.0 | 675 | 1.3018 | 0.3417 |
| 1.249 | 4.0 | 900 | 1.2839 | 0.3417 |
| 1.3057 | 5.0 | 1125 | 1.2676 | 0.345 |
| 1.2751 | 6.0 | 1350 | 1.2530 | 0.35 |
| 1.2676 | 7.0 | 1575 | 1.2399 | 0.3483 |
| 1.1849 | 8.0 | 1800 | 1.2280 | 0.355 |
| 1.2324 | 9.0 | 2025 | 1.2173 | 0.365 |
| 1.2387 | 10.0 | 2250 | 1.2077 | 0.365 |
| 1.1505 | 11.0 | 2475 | 1.1990 | 0.3767 |
| 1.2698 | 12.0 | 2700 | 1.1911 | 0.37 |
| 1.1755 | 13.0 | 2925 | 1.1841 | 0.3717 |
| 1.1872 | 14.0 | 3150 | 1.1777 | 0.3717 |
| 1.2051 | 15.0 | 3375 | 1.1720 | 0.3683 |
| 1.1902 | 16.0 | 3600 | 1.1669 | 0.3733 |
| 1.144 | 17.0 | 3825 | 1.1622 | 0.38 |
| 1.0991 | 18.0 | 4050 | 1.1579 | 0.375 |
| 1.1454 | 19.0 | 4275 | 1.1539 | 0.38 |
| 1.1739 | 20.0 | 4500 | 1.1503 | 0.3717 |
| 1.193 | 21.0 | 4725 | 1.1470 | 0.375 |
| 1.154 | 22.0 | 4950 | 1.1439 | 0.3783 |
| 1.1465 | 23.0 | 5175 | 1.1410 | 0.3733 |
| 1.1397 | 24.0 | 5400 | 1.1384 | 0.3733 |
| 1.1433 | 25.0 | 5625 | 1.1360 | 0.375 |
| 1.1492 | 26.0 | 5850 | 1.1337 | 0.3717 |
| 1.1616 | 27.0 | 6075 | 1.1316 | 0.37 |
| 1.137 | 28.0 | 6300 | 1.1297 | 0.3767 |
| 1.1705 | 29.0 | 6525 | 1.1279 | 0.3817 |
| 1.1514 | 30.0 | 6750 | 1.1262 | 0.3817 |
| 1.1348 | 31.0 | 6975 | 1.1247 | 0.385 |
| 1.152 | 32.0 | 7200 | 1.1232 | 0.3817 |
| 1.1108 | 33.0 | 7425 | 1.1219 | 0.385 |
| 1.1032 | 34.0 | 7650 | 1.1207 | 0.3867 |
| 1.0973 | 35.0 | 7875 | 1.1195 | 0.3833 |
| 1.1492 | 36.0 | 8100 | 1.1185 | 0.3867 |
| 1.1046 | 37.0 | 8325 | 1.1176 | 0.3883 |
| 1.1281 | 38.0 | 8550 | 1.1167 | 0.3883 |
| 1.0952 | 39.0 | 8775 | 1.1160 | 0.3917 |
| 1.0675 | 40.0 | 9000 | 1.1153 | 0.3883 |
| 1.1354 | 41.0 | 9225 | 1.1147 | 0.39 |
| 1.1352 | 42.0 | 9450 | 1.1142 | 0.3883 |
| 1.1158 | 43.0 | 9675 | 1.1138 | 0.39 |
| 1.1146 | 44.0 | 9900 | 1.1135 | 0.39 |
| 1.0831 | 45.0 | 10125 | 1.1132 | 0.39 |
| 1.1243 | 46.0 | 10350 | 1.1130 | 0.39 |
| 1.0703 | 47.0 | 10575 | 1.1128 | 0.3917 |
| 1.1356 | 48.0 | 10800 | 1.1127 | 0.3917 |
| 1.1249 | 49.0 | 11025 | 1.1127 | 0.3917 |
| 1.0942 | 50.0 | 11250 | 1.1127 | 0.3917 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2