hushem_5x_deit_small_rms_001_fold5
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: 1.2275
- Accuracy: 0.6585
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 |
|---|---|---|---|---|
| 1.7524 | 1.0 | 28 | 1.5298 | 0.2439 |
| 1.4312 | 2.0 | 56 | 1.4291 | 0.2683 |
| 1.3924 | 3.0 | 84 | 1.4059 | 0.2927 |
| 1.4173 | 4.0 | 112 | 1.3938 | 0.2683 |
| 1.3939 | 5.0 | 140 | 1.3790 | 0.2683 |
| 1.3863 | 6.0 | 168 | 1.4164 | 0.2439 |
| 1.3865 | 7.0 | 196 | 1.3790 | 0.2683 |
| 1.394 | 8.0 | 224 | 1.3790 | 0.2683 |
| 1.3883 | 9.0 | 252 | 1.4097 | 0.2683 |
| 1.3472 | 10.0 | 280 | 1.2478 | 0.4390 |
| 1.3905 | 11.0 | 308 | 1.2068 | 0.3902 |
| 1.1031 | 12.0 | 336 | 1.2038 | 0.4390 |
| 1.1503 | 13.0 | 364 | 1.0846 | 0.4634 |
| 1.2064 | 14.0 | 392 | 1.1395 | 0.4146 |
| 1.1249 | 15.0 | 420 | 1.1544 | 0.4146 |
| 1.1285 | 16.0 | 448 | 1.0714 | 0.4634 |
| 1.1149 | 17.0 | 476 | 0.9771 | 0.6098 |
| 1.0493 | 18.0 | 504 | 0.9974 | 0.4634 |
| 0.9938 | 19.0 | 532 | 0.9792 | 0.5366 |
| 1.0212 | 20.0 | 560 | 0.9949 | 0.5854 |
| 0.9943 | 21.0 | 588 | 1.0078 | 0.5366 |
| 1.0044 | 22.0 | 616 | 0.9007 | 0.5366 |
| 1.0661 | 23.0 | 644 | 1.2742 | 0.4878 |
| 0.9523 | 24.0 | 672 | 0.9851 | 0.6829 |
| 0.8733 | 25.0 | 700 | 0.9430 | 0.5854 |
| 0.8075 | 26.0 | 728 | 0.9660 | 0.6585 |
| 0.9128 | 27.0 | 756 | 0.9161 | 0.7561 |
| 0.8898 | 28.0 | 784 | 0.8767 | 0.7073 |
| 0.8051 | 29.0 | 812 | 0.8174 | 0.6829 |
| 0.8328 | 30.0 | 840 | 0.8077 | 0.6585 |
| 0.81 | 31.0 | 868 | 0.7911 | 0.6585 |
| 0.7372 | 32.0 | 896 | 1.0262 | 0.6585 |
| 0.7641 | 33.0 | 924 | 1.0698 | 0.5854 |
| 0.7745 | 34.0 | 952 | 0.8530 | 0.6829 |
| 0.7037 | 35.0 | 980 | 1.0106 | 0.6585 |
| 0.7449 | 36.0 | 1008 | 0.8975 | 0.7073 |
| 0.7391 | 37.0 | 1036 | 0.9607 | 0.6829 |
| 0.7447 | 38.0 | 1064 | 1.0096 | 0.6585 |
| 0.7043 | 39.0 | 1092 | 1.0986 | 0.7073 |
| 0.6379 | 40.0 | 1120 | 1.0787 | 0.6829 |
| 0.6476 | 41.0 | 1148 | 1.0057 | 0.6829 |
| 0.5799 | 42.0 | 1176 | 1.1714 | 0.6341 |
| 0.5954 | 43.0 | 1204 | 1.1356 | 0.6829 |
| 0.6189 | 44.0 | 1232 | 1.1609 | 0.6829 |
| 0.5672 | 45.0 | 1260 | 1.1726 | 0.6829 |
| 0.5115 | 46.0 | 1288 | 1.2388 | 0.6829 |
| 0.4522 | 47.0 | 1316 | 1.2273 | 0.6829 |
| 0.4728 | 48.0 | 1344 | 1.2290 | 0.6585 |
| 0.4195 | 49.0 | 1372 | 1.2275 | 0.6585 |
| 0.4871 | 50.0 | 1400 | 1.2275 | 0.6585 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0