hushem_5x_deit_tiny_adamax_00001_fold1
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: 1.3516
- Accuracy: 0.6444
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 |
|---|---|---|---|---|
| 1.3422 | 1.0 | 27 | 1.4246 | 0.2444 |
| 1.0091 | 2.0 | 54 | 1.3182 | 0.3333 |
| 0.8493 | 3.0 | 81 | 1.2550 | 0.3556 |
| 0.7491 | 4.0 | 108 | 1.2490 | 0.4 |
| 0.5567 | 5.0 | 135 | 1.2577 | 0.4444 |
| 0.4888 | 6.0 | 162 | 1.1346 | 0.4444 |
| 0.3998 | 7.0 | 189 | 1.0755 | 0.4889 |
| 0.3183 | 8.0 | 216 | 1.0806 | 0.5111 |
| 0.2348 | 9.0 | 243 | 1.0626 | 0.4667 |
| 0.1921 | 10.0 | 270 | 1.0210 | 0.5556 |
| 0.1327 | 11.0 | 297 | 1.0080 | 0.5556 |
| 0.0957 | 12.0 | 324 | 1.0134 | 0.5111 |
| 0.0834 | 13.0 | 351 | 1.0055 | 0.5556 |
| 0.0584 | 14.0 | 378 | 0.9707 | 0.5778 |
| 0.0407 | 15.0 | 405 | 0.9642 | 0.6 |
| 0.0299 | 16.0 | 432 | 0.9671 | 0.6222 |
| 0.0236 | 17.0 | 459 | 1.0493 | 0.6 |
| 0.0189 | 18.0 | 486 | 1.0428 | 0.6 |
| 0.0123 | 19.0 | 513 | 1.0847 | 0.6222 |
| 0.0089 | 20.0 | 540 | 1.0847 | 0.6667 |
| 0.0061 | 21.0 | 567 | 1.1226 | 0.6 |
| 0.0054 | 22.0 | 594 | 1.1899 | 0.6222 |
| 0.0037 | 23.0 | 621 | 1.2113 | 0.6444 |
| 0.0029 | 24.0 | 648 | 1.2131 | 0.6222 |
| 0.0025 | 25.0 | 675 | 1.2220 | 0.6222 |
| 0.0023 | 26.0 | 702 | 1.2400 | 0.6444 |
| 0.0022 | 27.0 | 729 | 1.2553 | 0.6444 |
| 0.0017 | 28.0 | 756 | 1.2656 | 0.6667 |
| 0.0018 | 29.0 | 783 | 1.2717 | 0.6444 |
| 0.0016 | 30.0 | 810 | 1.2861 | 0.6444 |
| 0.0014 | 31.0 | 837 | 1.2858 | 0.6444 |
| 0.0015 | 32.0 | 864 | 1.2948 | 0.6444 |
| 0.0012 | 33.0 | 891 | 1.3044 | 0.6444 |
| 0.0012 | 34.0 | 918 | 1.3113 | 0.6444 |
| 0.0011 | 35.0 | 945 | 1.3140 | 0.6444 |
| 0.0011 | 36.0 | 972 | 1.3218 | 0.6444 |
| 0.001 | 37.0 | 999 | 1.3254 | 0.6444 |
| 0.001 | 38.0 | 1026 | 1.3310 | 0.6444 |
| 0.0009 | 39.0 | 1053 | 1.3375 | 0.6444 |
| 0.0009 | 40.0 | 1080 | 1.3355 | 0.6444 |
| 0.0009 | 41.0 | 1107 | 1.3390 | 0.6444 |
| 0.0009 | 42.0 | 1134 | 1.3441 | 0.6444 |
| 0.0008 | 43.0 | 1161 | 1.3451 | 0.6444 |
| 0.0008 | 44.0 | 1188 | 1.3481 | 0.6444 |
| 0.0009 | 45.0 | 1215 | 1.3501 | 0.6444 |
| 0.0008 | 46.0 | 1242 | 1.3500 | 0.6444 |
| 0.0008 | 47.0 | 1269 | 1.3513 | 0.6444 |
| 0.0008 | 48.0 | 1296 | 1.3516 | 0.6444 |
| 0.0008 | 49.0 | 1323 | 1.3516 | 0.6444 |
| 0.0008 | 50.0 | 1350 | 1.3516 | 0.6444 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0