smids_1x_deit_tiny_sgd_0001_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: 0.8723
- Accuracy: 0.6023
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 |
|---|---|---|---|---|
| 1.2913 | 1.0 | 75 | 1.2725 | 0.3394 |
| 1.2246 | 2.0 | 150 | 1.2089 | 0.3444 |
| 1.1634 | 3.0 | 225 | 1.1639 | 0.3561 |
| 1.1405 | 4.0 | 300 | 1.1319 | 0.3661 |
| 1.1027 | 5.0 | 375 | 1.1096 | 0.3727 |
| 1.1314 | 6.0 | 450 | 1.0924 | 0.3960 |
| 1.1132 | 7.0 | 525 | 1.0784 | 0.4193 |
| 1.0729 | 8.0 | 600 | 1.0662 | 0.4359 |
| 1.064 | 9.0 | 675 | 1.0549 | 0.4526 |
| 1.0806 | 10.0 | 750 | 1.0450 | 0.4576 |
| 1.063 | 11.0 | 825 | 1.0356 | 0.4692 |
| 1.0489 | 12.0 | 900 | 1.0265 | 0.4792 |
| 1.0267 | 13.0 | 975 | 1.0180 | 0.4942 |
| 0.9878 | 14.0 | 1050 | 1.0096 | 0.5042 |
| 1.01 | 15.0 | 1125 | 1.0018 | 0.5058 |
| 0.9915 | 16.0 | 1200 | 0.9940 | 0.5092 |
| 0.9952 | 17.0 | 1275 | 0.9865 | 0.5158 |
| 1.0114 | 18.0 | 1350 | 0.9793 | 0.5258 |
| 1.0011 | 19.0 | 1425 | 0.9723 | 0.5308 |
| 0.9762 | 20.0 | 1500 | 0.9654 | 0.5358 |
| 1.0144 | 21.0 | 1575 | 0.9587 | 0.5408 |
| 0.9349 | 22.0 | 1650 | 0.9525 | 0.5507 |
| 0.9869 | 23.0 | 1725 | 0.9462 | 0.5557 |
| 0.9417 | 24.0 | 1800 | 0.9404 | 0.5591 |
| 0.9277 | 25.0 | 1875 | 0.9347 | 0.5591 |
| 0.9227 | 26.0 | 1950 | 0.9293 | 0.5707 |
| 0.9725 | 27.0 | 2025 | 0.9242 | 0.5674 |
| 0.9104 | 28.0 | 2100 | 0.9193 | 0.5691 |
| 0.9618 | 29.0 | 2175 | 0.9147 | 0.5774 |
| 0.8904 | 30.0 | 2250 | 0.9103 | 0.5824 |
| 0.9175 | 31.0 | 2325 | 0.9062 | 0.5840 |
| 0.916 | 32.0 | 2400 | 0.9024 | 0.5857 |
| 0.8843 | 33.0 | 2475 | 0.8990 | 0.5874 |
| 0.9346 | 34.0 | 2550 | 0.8956 | 0.5923 |
| 0.8711 | 35.0 | 2625 | 0.8924 | 0.5957 |
| 0.8808 | 36.0 | 2700 | 0.8897 | 0.5973 |
| 0.9043 | 37.0 | 2775 | 0.8870 | 0.5990 |
| 0.9738 | 38.0 | 2850 | 0.8847 | 0.5957 |
| 0.8643 | 39.0 | 2925 | 0.8826 | 0.5940 |
| 0.8918 | 40.0 | 3000 | 0.8806 | 0.5957 |
| 0.9135 | 41.0 | 3075 | 0.8789 | 0.5973 |
| 0.9187 | 42.0 | 3150 | 0.8773 | 0.5973 |
| 0.8721 | 43.0 | 3225 | 0.8760 | 0.5973 |
| 0.9047 | 44.0 | 3300 | 0.8749 | 0.5990 |
| 0.8511 | 45.0 | 3375 | 0.8740 | 0.6007 |
| 0.8649 | 46.0 | 3450 | 0.8732 | 0.6007 |
| 0.8917 | 47.0 | 3525 | 0.8727 | 0.6023 |
| 0.8901 | 48.0 | 3600 | 0.8724 | 0.6023 |
| 0.883 | 49.0 | 3675 | 0.8723 | 0.6023 |
| 0.8823 | 50.0 | 3750 | 0.8723 | 0.6023 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0