smids_3x_deit_tiny_sgd_00001_fold2
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: 1.1011
- Accuracy: 0.4027
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 |
|---|---|---|---|---|
| 1.2975 | 1.0 | 225 | 1.3313 | 0.3461 |
| 1.3535 | 2.0 | 450 | 1.3098 | 0.3411 |
| 1.364 | 3.0 | 675 | 1.2899 | 0.3411 |
| 1.2928 | 4.0 | 900 | 1.2721 | 0.3411 |
| 1.2627 | 5.0 | 1125 | 1.2559 | 0.3378 |
| 1.2073 | 6.0 | 1350 | 1.2414 | 0.3461 |
| 1.3184 | 7.0 | 1575 | 1.2280 | 0.3527 |
| 1.2058 | 8.0 | 1800 | 1.2162 | 0.3527 |
| 1.2305 | 9.0 | 2025 | 1.2057 | 0.3511 |
| 1.2453 | 10.0 | 2250 | 1.1960 | 0.3478 |
| 1.1822 | 11.0 | 2475 | 1.1875 | 0.3461 |
| 1.1856 | 12.0 | 2700 | 1.1797 | 0.3561 |
| 1.1979 | 13.0 | 2925 | 1.1728 | 0.3661 |
| 1.1589 | 14.0 | 3150 | 1.1665 | 0.3644 |
| 1.1625 | 15.0 | 3375 | 1.1608 | 0.3677 |
| 1.1751 | 16.0 | 3600 | 1.1557 | 0.3744 |
| 1.1846 | 17.0 | 3825 | 1.1510 | 0.3760 |
| 1.1541 | 18.0 | 4050 | 1.1466 | 0.3744 |
| 1.1807 | 19.0 | 4275 | 1.1426 | 0.3727 |
| 1.1744 | 20.0 | 4500 | 1.1389 | 0.3710 |
| 1.1694 | 21.0 | 4725 | 1.1356 | 0.3710 |
| 1.1819 | 22.0 | 4950 | 1.1325 | 0.3727 |
| 1.1574 | 23.0 | 5175 | 1.1297 | 0.3794 |
| 1.159 | 24.0 | 5400 | 1.1270 | 0.3760 |
| 1.1656 | 25.0 | 5625 | 1.1246 | 0.3760 |
| 1.1491 | 26.0 | 5850 | 1.1224 | 0.3777 |
| 1.1877 | 27.0 | 6075 | 1.1202 | 0.3760 |
| 1.1245 | 28.0 | 6300 | 1.1183 | 0.3810 |
| 1.1465 | 29.0 | 6525 | 1.1164 | 0.3877 |
| 1.0989 | 30.0 | 6750 | 1.1147 | 0.3910 |
| 1.1019 | 31.0 | 6975 | 1.1132 | 0.3927 |
| 1.1115 | 32.0 | 7200 | 1.1117 | 0.3927 |
| 1.1193 | 33.0 | 7425 | 1.1103 | 0.3943 |
| 1.1111 | 34.0 | 7650 | 1.1091 | 0.3960 |
| 1.1163 | 35.0 | 7875 | 1.1080 | 0.3977 |
| 1.1433 | 36.0 | 8100 | 1.1069 | 0.3993 |
| 1.0817 | 37.0 | 8325 | 1.1060 | 0.3993 |
| 1.1389 | 38.0 | 8550 | 1.1052 | 0.3993 |
| 1.1196 | 39.0 | 8775 | 1.1044 | 0.4027 |
| 1.1051 | 40.0 | 9000 | 1.1037 | 0.4043 |
| 1.1003 | 41.0 | 9225 | 1.1031 | 0.4027 |
| 1.1259 | 42.0 | 9450 | 1.1026 | 0.4027 |
| 1.1127 | 43.0 | 9675 | 1.1022 | 0.4027 |
| 1.1252 | 44.0 | 9900 | 1.1018 | 0.4010 |
| 1.0665 | 45.0 | 10125 | 1.1016 | 0.4027 |
| 1.1219 | 46.0 | 10350 | 1.1014 | 0.4027 |
| 1.1281 | 47.0 | 10575 | 1.1012 | 0.4027 |
| 1.0847 | 48.0 | 10800 | 1.1011 | 0.4027 |
| 1.1349 | 49.0 | 11025 | 1.1011 | 0.4027 |
| 1.1316 | 50.0 | 11250 | 1.1011 | 0.4027 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2