smids_1x_deit_small_sgd_0001_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: 0.7697
- Accuracy: 0.685
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: 0.0001
- 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 |
|---|---|---|---|---|
| 1.0786 | 1.0 | 75 | 1.0493 | 0.4483 |
| 1.0628 | 2.0 | 150 | 1.0330 | 0.4767 |
| 1.0285 | 3.0 | 225 | 1.0186 | 0.5233 |
| 0.9944 | 4.0 | 300 | 1.0062 | 0.53 |
| 1.0232 | 5.0 | 375 | 0.9947 | 0.5517 |
| 0.9927 | 6.0 | 450 | 0.9839 | 0.56 |
| 0.9942 | 7.0 | 525 | 0.9736 | 0.5767 |
| 0.9843 | 8.0 | 600 | 0.9634 | 0.595 |
| 0.9686 | 9.0 | 675 | 0.9535 | 0.6033 |
| 0.9669 | 10.0 | 750 | 0.9439 | 0.6067 |
| 0.9496 | 11.0 | 825 | 0.9345 | 0.6117 |
| 0.9424 | 12.0 | 900 | 0.9255 | 0.615 |
| 0.9379 | 13.0 | 975 | 0.9166 | 0.615 |
| 0.9246 | 14.0 | 1050 | 0.9079 | 0.6217 |
| 0.9261 | 15.0 | 1125 | 0.8998 | 0.63 |
| 0.8974 | 16.0 | 1200 | 0.8916 | 0.6333 |
| 0.9045 | 17.0 | 1275 | 0.8836 | 0.6367 |
| 0.8617 | 18.0 | 1350 | 0.8760 | 0.6417 |
| 0.885 | 19.0 | 1425 | 0.8688 | 0.6433 |
| 0.8736 | 20.0 | 1500 | 0.8617 | 0.6467 |
| 0.8843 | 21.0 | 1575 | 0.8551 | 0.6433 |
| 0.8472 | 22.0 | 1650 | 0.8488 | 0.6417 |
| 0.8796 | 23.0 | 1725 | 0.8428 | 0.6417 |
| 0.8784 | 24.0 | 1800 | 0.8370 | 0.6467 |
| 0.8408 | 25.0 | 1875 | 0.8316 | 0.65 |
| 0.8377 | 26.0 | 1950 | 0.8263 | 0.655 |
| 0.8101 | 27.0 | 2025 | 0.8213 | 0.6583 |
| 0.8334 | 28.0 | 2100 | 0.8166 | 0.66 |
| 0.8187 | 29.0 | 2175 | 0.8122 | 0.6567 |
| 0.8337 | 30.0 | 2250 | 0.8080 | 0.6583 |
| 0.8018 | 31.0 | 2325 | 0.8041 | 0.665 |
| 0.8384 | 32.0 | 2400 | 0.8003 | 0.67 |
| 0.813 | 33.0 | 2475 | 0.7968 | 0.6767 |
| 0.7997 | 34.0 | 2550 | 0.7936 | 0.6817 |
| 0.7882 | 35.0 | 2625 | 0.7905 | 0.6833 |
| 0.7651 | 36.0 | 2700 | 0.7878 | 0.6817 |
| 0.7706 | 37.0 | 2775 | 0.7852 | 0.6817 |
| 0.7916 | 38.0 | 2850 | 0.7828 | 0.6783 |
| 0.8116 | 39.0 | 2925 | 0.7807 | 0.6783 |
| 0.7662 | 40.0 | 3000 | 0.7787 | 0.6783 |
| 0.7857 | 41.0 | 3075 | 0.7769 | 0.6817 |
| 0.7862 | 42.0 | 3150 | 0.7753 | 0.6817 |
| 0.8172 | 43.0 | 3225 | 0.7740 | 0.685 |
| 0.7812 | 44.0 | 3300 | 0.7728 | 0.6867 |
| 0.803 | 45.0 | 3375 | 0.7718 | 0.685 |
| 0.7949 | 46.0 | 3450 | 0.7710 | 0.685 |
| 0.779 | 47.0 | 3525 | 0.7704 | 0.685 |
| 0.7941 | 48.0 | 3600 | 0.7700 | 0.685 |
| 0.7892 | 49.0 | 3675 | 0.7698 | 0.685 |
| 0.7766 | 50.0 | 3750 | 0.7697 | 0.685 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0