smids_1x_deit_tiny_rms_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: 1.8052
- Accuracy: 0.765
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 |
|---|---|---|---|---|
| 1.2535 | 1.0 | 75 | 0.9734 | 0.4717 |
| 1.0172 | 2.0 | 150 | 0.8857 | 0.5217 |
| 0.9205 | 3.0 | 225 | 0.8219 | 0.5633 |
| 0.8404 | 4.0 | 300 | 0.8833 | 0.54 |
| 0.8125 | 5.0 | 375 | 0.7752 | 0.615 |
| 0.8375 | 6.0 | 450 | 0.7791 | 0.6133 |
| 0.7706 | 7.0 | 525 | 0.7651 | 0.6433 |
| 0.6843 | 8.0 | 600 | 0.7674 | 0.6083 |
| 0.717 | 9.0 | 675 | 0.7318 | 0.655 |
| 0.6266 | 10.0 | 750 | 0.7160 | 0.6867 |
| 0.674 | 11.0 | 825 | 0.6761 | 0.69 |
| 0.6618 | 12.0 | 900 | 0.7236 | 0.6433 |
| 0.6204 | 13.0 | 975 | 0.7093 | 0.6733 |
| 0.6403 | 14.0 | 1050 | 0.6526 | 0.7133 |
| 0.5728 | 15.0 | 1125 | 0.7313 | 0.6617 |
| 0.5566 | 16.0 | 1200 | 0.6152 | 0.7317 |
| 0.5735 | 17.0 | 1275 | 0.6901 | 0.7083 |
| 0.6111 | 18.0 | 1350 | 0.6429 | 0.7317 |
| 0.6075 | 19.0 | 1425 | 0.6044 | 0.7533 |
| 0.5675 | 20.0 | 1500 | 0.5922 | 0.7633 |
| 0.4747 | 21.0 | 1575 | 0.6118 | 0.7483 |
| 0.5157 | 22.0 | 1650 | 0.6322 | 0.7383 |
| 0.4995 | 23.0 | 1725 | 0.6300 | 0.745 |
| 0.4632 | 24.0 | 1800 | 0.6076 | 0.74 |
| 0.4596 | 25.0 | 1875 | 0.6047 | 0.7733 |
| 0.4702 | 26.0 | 1950 | 0.6096 | 0.7633 |
| 0.5043 | 27.0 | 2025 | 0.6045 | 0.7567 |
| 0.5051 | 28.0 | 2100 | 0.5905 | 0.75 |
| 0.4664 | 29.0 | 2175 | 0.6085 | 0.7567 |
| 0.3949 | 30.0 | 2250 | 0.6634 | 0.76 |
| 0.3708 | 31.0 | 2325 | 0.6461 | 0.7667 |
| 0.3964 | 32.0 | 2400 | 0.6482 | 0.7617 |
| 0.3827 | 33.0 | 2475 | 0.6696 | 0.76 |
| 0.3422 | 34.0 | 2550 | 0.6799 | 0.765 |
| 0.3716 | 35.0 | 2625 | 0.7307 | 0.7767 |
| 0.3007 | 36.0 | 2700 | 0.7490 | 0.7583 |
| 0.2019 | 37.0 | 2775 | 0.8838 | 0.7533 |
| 0.232 | 38.0 | 2850 | 0.8738 | 0.76 |
| 0.221 | 39.0 | 2925 | 0.8842 | 0.7733 |
| 0.1875 | 40.0 | 3000 | 1.0078 | 0.7383 |
| 0.203 | 41.0 | 3075 | 1.0476 | 0.7567 |
| 0.1699 | 42.0 | 3150 | 1.0739 | 0.7567 |
| 0.171 | 43.0 | 3225 | 1.1644 | 0.7417 |
| 0.1205 | 44.0 | 3300 | 1.2501 | 0.7533 |
| 0.0811 | 45.0 | 3375 | 1.2967 | 0.755 |
| 0.0202 | 46.0 | 3450 | 1.5619 | 0.745 |
| 0.0237 | 47.0 | 3525 | 1.5862 | 0.7617 |
| 0.0127 | 48.0 | 3600 | 1.6631 | 0.7667 |
| 0.0204 | 49.0 | 3675 | 1.7536 | 0.7667 |
| 0.0042 | 50.0 | 3750 | 1.8052 | 0.765 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0