hushem_40x_deit_tiny_adamax_001_fold5
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: 1.0843
- Accuracy: 0.8293
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 |
|---|---|---|---|---|
| 0.2361 | 1.0 | 220 | 0.7775 | 0.7805 |
| 0.2436 | 2.0 | 440 | 1.3306 | 0.7561 |
| 0.1462 | 3.0 | 660 | 0.9347 | 0.7805 |
| 0.071 | 4.0 | 880 | 0.5663 | 0.8537 |
| 0.1006 | 5.0 | 1100 | 0.2859 | 0.8780 |
| 0.1238 | 6.0 | 1320 | 0.9526 | 0.8293 |
| 0.0201 | 7.0 | 1540 | 1.1774 | 0.8049 |
| 0.0017 | 8.0 | 1760 | 1.4587 | 0.7561 |
| 0.0368 | 9.0 | 1980 | 0.7868 | 0.8537 |
| 0.0002 | 10.0 | 2200 | 0.8716 | 0.8780 |
| 0.0421 | 11.0 | 2420 | 0.9525 | 0.8049 |
| 0.0055 | 12.0 | 2640 | 1.5979 | 0.7805 |
| 0.0103 | 13.0 | 2860 | 0.4608 | 0.9024 |
| 0.0 | 14.0 | 3080 | 1.1806 | 0.8049 |
| 0.0 | 15.0 | 3300 | 1.1203 | 0.8293 |
| 0.0 | 16.0 | 3520 | 1.1285 | 0.8293 |
| 0.0 | 17.0 | 3740 | 1.1228 | 0.8293 |
| 0.0 | 18.0 | 3960 | 1.1188 | 0.8293 |
| 0.0 | 19.0 | 4180 | 1.1166 | 0.8293 |
| 0.0 | 20.0 | 4400 | 1.1122 | 0.8293 |
| 0.0 | 21.0 | 4620 | 1.1096 | 0.8293 |
| 0.0 | 22.0 | 4840 | 1.1085 | 0.8293 |
| 0.0 | 23.0 | 5060 | 1.1063 | 0.8293 |
| 0.0 | 24.0 | 5280 | 1.1040 | 0.8293 |
| 0.0 | 25.0 | 5500 | 1.1028 | 0.8293 |
| 0.0 | 26.0 | 5720 | 1.0996 | 0.8293 |
| 0.0 | 27.0 | 5940 | 1.0984 | 0.8293 |
| 0.0 | 28.0 | 6160 | 1.0966 | 0.8293 |
| 0.0 | 29.0 | 6380 | 1.0939 | 0.8293 |
| 0.0 | 30.0 | 6600 | 1.0930 | 0.8293 |
| 0.0 | 31.0 | 6820 | 1.0903 | 0.8293 |
| 0.0 | 32.0 | 7040 | 1.0890 | 0.8293 |
| 0.0 | 33.0 | 7260 | 1.0876 | 0.8293 |
| 0.0 | 34.0 | 7480 | 1.0855 | 0.8293 |
| 0.0 | 35.0 | 7700 | 1.0853 | 0.8293 |
| 0.0 | 36.0 | 7920 | 1.0829 | 0.8293 |
| 0.0 | 37.0 | 8140 | 1.0834 | 0.8293 |
| 0.0 | 38.0 | 8360 | 1.0821 | 0.8293 |
| 0.0 | 39.0 | 8580 | 1.0819 | 0.8293 |
| 0.0 | 40.0 | 8800 | 1.0819 | 0.8293 |
| 0.0 | 41.0 | 9020 | 1.0821 | 0.8293 |
| 0.0 | 42.0 | 9240 | 1.0825 | 0.8293 |
| 0.0 | 43.0 | 9460 | 1.0825 | 0.8293 |
| 0.0 | 44.0 | 9680 | 1.0818 | 0.8293 |
| 0.0 | 45.0 | 9900 | 1.0822 | 0.8293 |
| 0.0 | 46.0 | 10120 | 1.0832 | 0.8293 |
| 0.0 | 47.0 | 10340 | 1.0843 | 0.8293 |
| 0.0 | 48.0 | 10560 | 1.0840 | 0.8293 |
| 0.0 | 49.0 | 10780 | 1.0843 | 0.8293 |
| 0.0 | 50.0 | 11000 | 1.0843 | 0.8293 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2