smids_3x_deit_small_sgd_001_fold1
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: 0.3034
- Accuracy: 0.8765
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 |
|---|---|---|---|---|
| 0.874 | 1.0 | 226 | 0.8422 | 0.6477 |
| 0.6808 | 2.0 | 452 | 0.6564 | 0.7362 |
| 0.5184 | 3.0 | 678 | 0.5592 | 0.7780 |
| 0.5173 | 4.0 | 904 | 0.5087 | 0.7813 |
| 0.5113 | 5.0 | 1130 | 0.4746 | 0.8047 |
| 0.4469 | 6.0 | 1356 | 0.4445 | 0.8214 |
| 0.3731 | 7.0 | 1582 | 0.4247 | 0.8364 |
| 0.413 | 8.0 | 1808 | 0.4105 | 0.8397 |
| 0.369 | 9.0 | 2034 | 0.3943 | 0.8464 |
| 0.3882 | 10.0 | 2260 | 0.3826 | 0.8548 |
| 0.3525 | 11.0 | 2486 | 0.3772 | 0.8648 |
| 0.3235 | 12.0 | 2712 | 0.3644 | 0.8598 |
| 0.2669 | 13.0 | 2938 | 0.3578 | 0.8715 |
| 0.2463 | 14.0 | 3164 | 0.3531 | 0.8681 |
| 0.3167 | 15.0 | 3390 | 0.3484 | 0.8698 |
| 0.3258 | 16.0 | 3616 | 0.3417 | 0.8715 |
| 0.2311 | 17.0 | 3842 | 0.3383 | 0.8748 |
| 0.2669 | 18.0 | 4068 | 0.3359 | 0.8748 |
| 0.2828 | 19.0 | 4294 | 0.3332 | 0.8731 |
| 0.2409 | 20.0 | 4520 | 0.3289 | 0.8731 |
| 0.3064 | 21.0 | 4746 | 0.3277 | 0.8715 |
| 0.2918 | 22.0 | 4972 | 0.3274 | 0.8731 |
| 0.3068 | 23.0 | 5198 | 0.3218 | 0.8748 |
| 0.2544 | 24.0 | 5424 | 0.3201 | 0.8715 |
| 0.2207 | 25.0 | 5650 | 0.3191 | 0.8781 |
| 0.2923 | 26.0 | 5876 | 0.3153 | 0.8748 |
| 0.2033 | 27.0 | 6102 | 0.3156 | 0.8765 |
| 0.2492 | 28.0 | 6328 | 0.3128 | 0.8731 |
| 0.2136 | 29.0 | 6554 | 0.3157 | 0.8715 |
| 0.2344 | 30.0 | 6780 | 0.3143 | 0.8765 |
| 0.2394 | 31.0 | 7006 | 0.3111 | 0.8781 |
| 0.2259 | 32.0 | 7232 | 0.3105 | 0.8781 |
| 0.2144 | 33.0 | 7458 | 0.3085 | 0.8781 |
| 0.1831 | 34.0 | 7684 | 0.3101 | 0.8748 |
| 0.2542 | 35.0 | 7910 | 0.3071 | 0.8781 |
| 0.211 | 36.0 | 8136 | 0.3058 | 0.8781 |
| 0.2498 | 37.0 | 8362 | 0.3047 | 0.8815 |
| 0.2414 | 38.0 | 8588 | 0.3058 | 0.8765 |
| 0.1832 | 39.0 | 8814 | 0.3053 | 0.8765 |
| 0.1872 | 40.0 | 9040 | 0.3039 | 0.8781 |
| 0.1641 | 41.0 | 9266 | 0.3048 | 0.8765 |
| 0.244 | 42.0 | 9492 | 0.3047 | 0.8748 |
| 0.1877 | 43.0 | 9718 | 0.3040 | 0.8765 |
| 0.1335 | 44.0 | 9944 | 0.3029 | 0.8781 |
| 0.2501 | 45.0 | 10170 | 0.3036 | 0.8781 |
| 0.2161 | 46.0 | 10396 | 0.3032 | 0.8765 |
| 0.222 | 47.0 | 10622 | 0.3033 | 0.8781 |
| 0.2279 | 48.0 | 10848 | 0.3033 | 0.8765 |
| 0.1815 | 49.0 | 11074 | 0.3033 | 0.8765 |
| 0.1776 | 50.0 | 11300 | 0.3034 | 0.8765 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2