smids_1x_deit_small_rms_0001_fold5
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: 1.2122
- Accuracy: 0.7483
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 |
|---|---|---|---|---|
| 1.0524 | 1.0 | 75 | 0.9597 | 0.445 |
| 1.1247 | 2.0 | 150 | 1.1111 | 0.3367 |
| 0.979 | 3.0 | 225 | 0.9077 | 0.5 |
| 0.8898 | 4.0 | 300 | 0.8740 | 0.52 |
| 0.8714 | 5.0 | 375 | 0.9443 | 0.4433 |
| 0.8755 | 6.0 | 450 | 0.7908 | 0.5917 |
| 0.8257 | 7.0 | 525 | 0.8028 | 0.5817 |
| 0.7602 | 8.0 | 600 | 0.8435 | 0.605 |
| 0.7994 | 9.0 | 675 | 0.7977 | 0.6117 |
| 0.7424 | 10.0 | 750 | 0.7850 | 0.6117 |
| 0.8101 | 11.0 | 825 | 0.7616 | 0.6233 |
| 0.7712 | 12.0 | 900 | 0.7668 | 0.6367 |
| 0.7209 | 13.0 | 975 | 0.8101 | 0.62 |
| 0.7215 | 14.0 | 1050 | 0.7936 | 0.62 |
| 0.7097 | 15.0 | 1125 | 0.7953 | 0.61 |
| 0.7072 | 16.0 | 1200 | 0.7924 | 0.6317 |
| 0.7074 | 17.0 | 1275 | 0.7452 | 0.6667 |
| 0.6856 | 18.0 | 1350 | 0.7477 | 0.6717 |
| 0.6768 | 19.0 | 1425 | 0.7216 | 0.6783 |
| 0.6919 | 20.0 | 1500 | 0.7445 | 0.68 |
| 0.6145 | 21.0 | 1575 | 0.7497 | 0.6533 |
| 0.5852 | 22.0 | 1650 | 0.7462 | 0.7083 |
| 0.625 | 23.0 | 1725 | 0.7496 | 0.675 |
| 0.549 | 24.0 | 1800 | 0.7315 | 0.7067 |
| 0.5773 | 25.0 | 1875 | 0.7055 | 0.7033 |
| 0.5746 | 26.0 | 1950 | 0.6982 | 0.7283 |
| 0.5717 | 27.0 | 2025 | 0.7187 | 0.705 |
| 0.5927 | 28.0 | 2100 | 0.6996 | 0.7183 |
| 0.5713 | 29.0 | 2175 | 0.6989 | 0.7217 |
| 0.5709 | 30.0 | 2250 | 0.7204 | 0.7267 |
| 0.5164 | 31.0 | 2325 | 0.7778 | 0.705 |
| 0.5059 | 32.0 | 2400 | 0.7021 | 0.73 |
| 0.5725 | 33.0 | 2475 | 0.6873 | 0.735 |
| 0.4839 | 34.0 | 2550 | 0.6931 | 0.745 |
| 0.4617 | 35.0 | 2625 | 0.7517 | 0.75 |
| 0.4294 | 36.0 | 2700 | 0.8099 | 0.7533 |
| 0.3749 | 37.0 | 2775 | 0.7255 | 0.75 |
| 0.4163 | 38.0 | 2850 | 0.7476 | 0.7533 |
| 0.3565 | 39.0 | 2925 | 0.8354 | 0.735 |
| 0.382 | 40.0 | 3000 | 0.8201 | 0.7467 |
| 0.3261 | 41.0 | 3075 | 0.8167 | 0.7567 |
| 0.4372 | 42.0 | 3150 | 0.8428 | 0.7267 |
| 0.3484 | 43.0 | 3225 | 0.8996 | 0.74 |
| 0.3261 | 44.0 | 3300 | 0.9207 | 0.735 |
| 0.2963 | 45.0 | 3375 | 1.0220 | 0.7283 |
| 0.2143 | 46.0 | 3450 | 0.9860 | 0.755 |
| 0.2551 | 47.0 | 3525 | 1.1473 | 0.7333 |
| 0.1675 | 48.0 | 3600 | 1.1351 | 0.735 |
| 0.1431 | 49.0 | 3675 | 1.1685 | 0.75 |
| 0.1393 | 50.0 | 3750 | 1.2122 | 0.7483 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0