smids_3x_deit_small_sgd_00001_fold3
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: 1.0287
- Accuracy: 0.4617
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 |
|---|---|---|---|---|
| 1.0656 | 1.0 | 225 | 1.0860 | 0.385 |
| 1.0566 | 2.0 | 450 | 1.0832 | 0.385 |
| 1.0608 | 3.0 | 675 | 1.0806 | 0.385 |
| 1.0487 | 4.0 | 900 | 1.0780 | 0.3883 |
| 1.0631 | 5.0 | 1125 | 1.0755 | 0.39 |
| 1.0618 | 6.0 | 1350 | 1.0731 | 0.395 |
| 1.0528 | 7.0 | 1575 | 1.0708 | 0.3967 |
| 1.0523 | 8.0 | 1800 | 1.0686 | 0.3967 |
| 1.0663 | 9.0 | 2025 | 1.0664 | 0.3983 |
| 1.0433 | 10.0 | 2250 | 1.0643 | 0.405 |
| 1.0514 | 11.0 | 2475 | 1.0623 | 0.4067 |
| 1.0454 | 12.0 | 2700 | 1.0603 | 0.4083 |
| 1.0616 | 13.0 | 2925 | 1.0585 | 0.41 |
| 1.031 | 14.0 | 3150 | 1.0567 | 0.415 |
| 1.0471 | 15.0 | 3375 | 1.0550 | 0.42 |
| 1.0587 | 16.0 | 3600 | 1.0533 | 0.42 |
| 1.0376 | 17.0 | 3825 | 1.0517 | 0.4233 |
| 1.0297 | 18.0 | 4050 | 1.0502 | 0.4267 |
| 1.0331 | 19.0 | 4275 | 1.0487 | 0.435 |
| 1.0488 | 20.0 | 4500 | 1.0473 | 0.4367 |
| 1.0355 | 21.0 | 4725 | 1.0459 | 0.4367 |
| 1.0375 | 22.0 | 4950 | 1.0446 | 0.4367 |
| 1.0233 | 23.0 | 5175 | 1.0434 | 0.4367 |
| 1.0207 | 24.0 | 5400 | 1.0422 | 0.44 |
| 1.0243 | 25.0 | 5625 | 1.0410 | 0.445 |
| 1.0105 | 26.0 | 5850 | 1.0400 | 0.4467 |
| 1.019 | 27.0 | 6075 | 1.0389 | 0.4467 |
| 1.0208 | 28.0 | 6300 | 1.0379 | 0.4467 |
| 1.0103 | 29.0 | 6525 | 1.0370 | 0.4483 |
| 1.0126 | 30.0 | 6750 | 1.0362 | 0.4517 |
| 1.0069 | 31.0 | 6975 | 1.0354 | 0.4533 |
| 1.0415 | 32.0 | 7200 | 1.0346 | 0.455 |
| 1.0107 | 33.0 | 7425 | 1.0339 | 0.4567 |
| 1.013 | 34.0 | 7650 | 1.0332 | 0.4583 |
| 0.9989 | 35.0 | 7875 | 1.0326 | 0.4583 |
| 0.9864 | 36.0 | 8100 | 1.0320 | 0.4583 |
| 1.0065 | 37.0 | 8325 | 1.0315 | 0.4583 |
| 1.0147 | 38.0 | 8550 | 1.0310 | 0.46 |
| 1.0279 | 39.0 | 8775 | 1.0306 | 0.4583 |
| 1.0198 | 40.0 | 9000 | 1.0302 | 0.4583 |
| 1.0298 | 41.0 | 9225 | 1.0299 | 0.4583 |
| 1.0219 | 42.0 | 9450 | 1.0296 | 0.4583 |
| 1.0277 | 43.0 | 9675 | 1.0293 | 0.4583 |
| 1.0097 | 44.0 | 9900 | 1.0291 | 0.4583 |
| 1.0316 | 45.0 | 10125 | 1.0290 | 0.4583 |
| 1.0133 | 46.0 | 10350 | 1.0288 | 0.46 |
| 0.9895 | 47.0 | 10575 | 1.0288 | 0.4617 |
| 1.0235 | 48.0 | 10800 | 1.0287 | 0.4617 |
| 1.0129 | 49.0 | 11025 | 1.0287 | 0.4617 |
| 0.9906 | 50.0 | 11250 | 1.0287 | 0.4617 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2