smids_1x_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: 0.3679
- Accuracy: 0.8617
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 |
|---|---|---|---|---|
| 0.9928 | 1.0 | 75 | 0.9393 | 0.6083 |
| 0.8712 | 2.0 | 150 | 0.8295 | 0.6433 |
| 0.7828 | 3.0 | 225 | 0.7388 | 0.7083 |
| 0.6696 | 4.0 | 300 | 0.6714 | 0.7567 |
| 0.6657 | 5.0 | 375 | 0.6202 | 0.7767 |
| 0.5692 | 6.0 | 450 | 0.5823 | 0.7883 |
| 0.568 | 7.0 | 525 | 0.5504 | 0.795 |
| 0.5196 | 8.0 | 600 | 0.5262 | 0.805 |
| 0.5314 | 9.0 | 675 | 0.5056 | 0.8167 |
| 0.4867 | 10.0 | 750 | 0.4885 | 0.8183 |
| 0.4734 | 11.0 | 825 | 0.4740 | 0.8183 |
| 0.4623 | 12.0 | 900 | 0.4619 | 0.8267 |
| 0.4683 | 13.0 | 975 | 0.4510 | 0.835 |
| 0.4153 | 14.0 | 1050 | 0.4430 | 0.8317 |
| 0.3992 | 15.0 | 1125 | 0.4360 | 0.8333 |
| 0.3763 | 16.0 | 1200 | 0.4269 | 0.845 |
| 0.3576 | 17.0 | 1275 | 0.4206 | 0.85 |
| 0.3638 | 18.0 | 1350 | 0.4146 | 0.845 |
| 0.3788 | 19.0 | 1425 | 0.4098 | 0.8433 |
| 0.333 | 20.0 | 1500 | 0.4060 | 0.85 |
| 0.3671 | 21.0 | 1575 | 0.4016 | 0.85 |
| 0.3178 | 22.0 | 1650 | 0.3983 | 0.8533 |
| 0.3335 | 23.0 | 1725 | 0.3950 | 0.8467 |
| 0.3527 | 24.0 | 1800 | 0.3923 | 0.8533 |
| 0.3211 | 25.0 | 1875 | 0.3897 | 0.8483 |
| 0.3209 | 26.0 | 1950 | 0.3876 | 0.855 |
| 0.2907 | 27.0 | 2025 | 0.3854 | 0.8517 |
| 0.3294 | 28.0 | 2100 | 0.3833 | 0.86 |
| 0.2805 | 29.0 | 2175 | 0.3818 | 0.8533 |
| 0.3183 | 30.0 | 2250 | 0.3799 | 0.855 |
| 0.2622 | 31.0 | 2325 | 0.3784 | 0.86 |
| 0.3165 | 32.0 | 2400 | 0.3771 | 0.86 |
| 0.2898 | 33.0 | 2475 | 0.3761 | 0.8617 |
| 0.2776 | 34.0 | 2550 | 0.3750 | 0.8617 |
| 0.2847 | 35.0 | 2625 | 0.3739 | 0.8633 |
| 0.2631 | 36.0 | 2700 | 0.3731 | 0.86 |
| 0.2584 | 37.0 | 2775 | 0.3723 | 0.8583 |
| 0.2889 | 38.0 | 2850 | 0.3718 | 0.8617 |
| 0.3213 | 39.0 | 2925 | 0.3709 | 0.8617 |
| 0.2873 | 40.0 | 3000 | 0.3704 | 0.865 |
| 0.284 | 41.0 | 3075 | 0.3700 | 0.865 |
| 0.2939 | 42.0 | 3150 | 0.3694 | 0.8633 |
| 0.3121 | 43.0 | 3225 | 0.3692 | 0.8633 |
| 0.2744 | 44.0 | 3300 | 0.3688 | 0.8617 |
| 0.2881 | 45.0 | 3375 | 0.3686 | 0.8617 |
| 0.302 | 46.0 | 3450 | 0.3683 | 0.8617 |
| 0.2645 | 47.0 | 3525 | 0.3682 | 0.8617 |
| 0.307 | 48.0 | 3600 | 0.3681 | 0.8617 |
| 0.2733 | 49.0 | 3675 | 0.3680 | 0.8617 |
| 0.2676 | 50.0 | 3750 | 0.3679 | 0.8617 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0