hushem_40x_deit_small_sgd_00001_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: 1.4814
- Accuracy: 0.2619
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 |
|---|---|---|---|---|
| 1.9934 | 1.0 | 219 | 1.6922 | 0.2619 |
| 1.9676 | 2.0 | 438 | 1.6780 | 0.2619 |
| 1.8623 | 3.0 | 657 | 1.6644 | 0.2619 |
| 1.8471 | 4.0 | 876 | 1.6514 | 0.2619 |
| 1.8032 | 5.0 | 1095 | 1.6391 | 0.2619 |
| 1.7347 | 6.0 | 1314 | 1.6275 | 0.2619 |
| 1.7421 | 7.0 | 1533 | 1.6165 | 0.2619 |
| 1.7487 | 8.0 | 1752 | 1.6060 | 0.2857 |
| 1.7488 | 9.0 | 1971 | 1.5962 | 0.2857 |
| 1.6907 | 10.0 | 2190 | 1.5869 | 0.2857 |
| 1.7571 | 11.0 | 2409 | 1.5783 | 0.2857 |
| 1.7123 | 12.0 | 2628 | 1.5703 | 0.3095 |
| 1.6705 | 13.0 | 2847 | 1.5628 | 0.3095 |
| 1.6677 | 14.0 | 3066 | 1.5558 | 0.3095 |
| 1.7011 | 15.0 | 3285 | 1.5494 | 0.3095 |
| 1.6459 | 16.0 | 3504 | 1.5434 | 0.3095 |
| 1.6424 | 17.0 | 3723 | 1.5379 | 0.2857 |
| 1.666 | 18.0 | 3942 | 1.5328 | 0.2857 |
| 1.5793 | 19.0 | 4161 | 1.5282 | 0.2857 |
| 1.612 | 20.0 | 4380 | 1.5239 | 0.2857 |
| 1.6378 | 21.0 | 4599 | 1.5199 | 0.2857 |
| 1.603 | 22.0 | 4818 | 1.5163 | 0.2857 |
| 1.5587 | 23.0 | 5037 | 1.5130 | 0.2619 |
| 1.5451 | 24.0 | 5256 | 1.5099 | 0.2619 |
| 1.6036 | 25.0 | 5475 | 1.5071 | 0.2619 |
| 1.5234 | 26.0 | 5694 | 1.5045 | 0.2857 |
| 1.5523 | 27.0 | 5913 | 1.5021 | 0.2857 |
| 1.5056 | 28.0 | 6132 | 1.4999 | 0.2857 |
| 1.5455 | 29.0 | 6351 | 1.4979 | 0.2857 |
| 1.5427 | 30.0 | 6570 | 1.4960 | 0.2857 |
| 1.566 | 31.0 | 6789 | 1.4943 | 0.2857 |
| 1.5088 | 32.0 | 7008 | 1.4927 | 0.2857 |
| 1.524 | 33.0 | 7227 | 1.4913 | 0.2857 |
| 1.5159 | 34.0 | 7446 | 1.4900 | 0.2857 |
| 1.4946 | 35.0 | 7665 | 1.4888 | 0.2619 |
| 1.5331 | 36.0 | 7884 | 1.4877 | 0.2619 |
| 1.5143 | 37.0 | 8103 | 1.4867 | 0.2619 |
| 1.5074 | 38.0 | 8322 | 1.4858 | 0.2619 |
| 1.4853 | 39.0 | 8541 | 1.4850 | 0.2619 |
| 1.5395 | 40.0 | 8760 | 1.4843 | 0.2619 |
| 1.4854 | 41.0 | 8979 | 1.4837 | 0.2619 |
| 1.4835 | 42.0 | 9198 | 1.4832 | 0.2619 |
| 1.5069 | 43.0 | 9417 | 1.4827 | 0.2619 |
| 1.4921 | 44.0 | 9636 | 1.4823 | 0.2619 |
| 1.4639 | 45.0 | 9855 | 1.4820 | 0.2619 |
| 1.511 | 46.0 | 10074 | 1.4818 | 0.2619 |
| 1.4986 | 47.0 | 10293 | 1.4816 | 0.2619 |
| 1.4648 | 48.0 | 10512 | 1.4815 | 0.2619 |
| 1.5244 | 49.0 | 10731 | 1.4814 | 0.2619 |
| 1.5102 | 50.0 | 10950 | 1.4814 | 0.2619 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2