smids_1x_deit_small_sgd_001_fold2
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: 0.3804
- Accuracy: 0.8453
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 |
|---|---|---|---|---|
| 0.9829 | 1.0 | 75 | 0.9415 | 0.6140 |
| 0.8706 | 2.0 | 150 | 0.8332 | 0.6855 |
| 0.7459 | 3.0 | 225 | 0.7472 | 0.7105 |
| 0.699 | 4.0 | 300 | 0.6837 | 0.7404 |
| 0.6391 | 5.0 | 375 | 0.6363 | 0.7587 |
| 0.5631 | 6.0 | 450 | 0.5989 | 0.7687 |
| 0.5887 | 7.0 | 525 | 0.5694 | 0.7820 |
| 0.5519 | 8.0 | 600 | 0.5464 | 0.7887 |
| 0.4995 | 9.0 | 675 | 0.5237 | 0.7970 |
| 0.5219 | 10.0 | 750 | 0.5075 | 0.8053 |
| 0.4633 | 11.0 | 825 | 0.4946 | 0.8070 |
| 0.4238 | 12.0 | 900 | 0.4804 | 0.8220 |
| 0.4323 | 13.0 | 975 | 0.4699 | 0.8253 |
| 0.358 | 14.0 | 1050 | 0.4601 | 0.8253 |
| 0.3922 | 15.0 | 1125 | 0.4517 | 0.8270 |
| 0.412 | 16.0 | 1200 | 0.4446 | 0.8253 |
| 0.3525 | 17.0 | 1275 | 0.4384 | 0.8236 |
| 0.3851 | 18.0 | 1350 | 0.4333 | 0.8253 |
| 0.4196 | 19.0 | 1425 | 0.4291 | 0.8353 |
| 0.3588 | 20.0 | 1500 | 0.4237 | 0.8286 |
| 0.4139 | 21.0 | 1575 | 0.4192 | 0.8353 |
| 0.3218 | 22.0 | 1650 | 0.4155 | 0.8369 |
| 0.3303 | 23.0 | 1725 | 0.4117 | 0.8419 |
| 0.3481 | 24.0 | 1800 | 0.4077 | 0.8419 |
| 0.3241 | 25.0 | 1875 | 0.4056 | 0.8403 |
| 0.3326 | 26.0 | 1950 | 0.4042 | 0.8453 |
| 0.3492 | 27.0 | 2025 | 0.4005 | 0.8453 |
| 0.2987 | 28.0 | 2100 | 0.3985 | 0.8419 |
| 0.3361 | 29.0 | 2175 | 0.3960 | 0.8453 |
| 0.2986 | 30.0 | 2250 | 0.3933 | 0.8469 |
| 0.2629 | 31.0 | 2325 | 0.3933 | 0.8469 |
| 0.3171 | 32.0 | 2400 | 0.3920 | 0.8453 |
| 0.2746 | 33.0 | 2475 | 0.3904 | 0.8453 |
| 0.2943 | 34.0 | 2550 | 0.3897 | 0.8453 |
| 0.2828 | 35.0 | 2625 | 0.3874 | 0.8453 |
| 0.2865 | 36.0 | 2700 | 0.3865 | 0.8453 |
| 0.2715 | 37.0 | 2775 | 0.3856 | 0.8453 |
| 0.3146 | 38.0 | 2850 | 0.3843 | 0.8453 |
| 0.2703 | 39.0 | 2925 | 0.3840 | 0.8453 |
| 0.2829 | 40.0 | 3000 | 0.3831 | 0.8453 |
| 0.2957 | 41.0 | 3075 | 0.3831 | 0.8453 |
| 0.29 | 42.0 | 3150 | 0.3825 | 0.8469 |
| 0.2671 | 43.0 | 3225 | 0.3820 | 0.8469 |
| 0.262 | 44.0 | 3300 | 0.3814 | 0.8453 |
| 0.2814 | 45.0 | 3375 | 0.3812 | 0.8436 |
| 0.272 | 46.0 | 3450 | 0.3810 | 0.8436 |
| 0.2705 | 47.0 | 3525 | 0.3806 | 0.8453 |
| 0.2571 | 48.0 | 3600 | 0.3805 | 0.8453 |
| 0.267 | 49.0 | 3675 | 0.3804 | 0.8453 |
| 0.2797 | 50.0 | 3750 | 0.3804 | 0.8453 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0