smids_5x_deit_small_adamax_0001_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: 1.1317
- Accuracy: 0.8918
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: 0.0001
- 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 |
|---|---|---|---|---|
| 0.2268 | 1.0 | 750 | 0.2931 | 0.8885 |
| 0.0856 | 2.0 | 1500 | 0.3354 | 0.9018 |
| 0.0454 | 3.0 | 2250 | 0.4505 | 0.8902 |
| 0.0358 | 4.0 | 3000 | 0.7256 | 0.8918 |
| 0.0161 | 5.0 | 3750 | 0.7228 | 0.8918 |
| 0.0143 | 6.0 | 4500 | 0.7580 | 0.8885 |
| 0.0111 | 7.0 | 5250 | 0.7663 | 0.9052 |
| 0.003 | 8.0 | 6000 | 0.8086 | 0.8852 |
| 0.0012 | 9.0 | 6750 | 0.7843 | 0.8968 |
| 0.0007 | 10.0 | 7500 | 0.8932 | 0.8935 |
| 0.0001 | 11.0 | 8250 | 0.9695 | 0.8902 |
| 0.0 | 12.0 | 9000 | 0.9681 | 0.8968 |
| 0.0 | 13.0 | 9750 | 0.9905 | 0.8869 |
| 0.0 | 14.0 | 10500 | 1.0234 | 0.8819 |
| 0.0001 | 15.0 | 11250 | 0.9740 | 0.9002 |
| 0.0354 | 16.0 | 12000 | 0.9561 | 0.9018 |
| 0.0 | 17.0 | 12750 | 0.9170 | 0.8902 |
| 0.0 | 18.0 | 13500 | 0.8495 | 0.9035 |
| 0.0 | 19.0 | 14250 | 0.9851 | 0.8852 |
| 0.0 | 20.0 | 15000 | 0.9671 | 0.8952 |
| 0.0 | 21.0 | 15750 | 1.0261 | 0.8885 |
| 0.004 | 22.0 | 16500 | 1.0404 | 0.8918 |
| 0.0023 | 23.0 | 17250 | 1.0326 | 0.8952 |
| 0.0 | 24.0 | 18000 | 1.0484 | 0.8852 |
| 0.0 | 25.0 | 18750 | 1.0611 | 0.8902 |
| 0.0 | 26.0 | 19500 | 1.0089 | 0.8985 |
| 0.0058 | 27.0 | 20250 | 1.0656 | 0.8935 |
| 0.0 | 28.0 | 21000 | 1.1306 | 0.8802 |
| 0.0 | 29.0 | 21750 | 1.0444 | 0.8985 |
| 0.0 | 30.0 | 22500 | 1.1060 | 0.8835 |
| 0.0 | 31.0 | 23250 | 1.0495 | 0.8935 |
| 0.0 | 32.0 | 24000 | 1.0881 | 0.8902 |
| 0.0 | 33.0 | 24750 | 1.0585 | 0.8968 |
| 0.0 | 34.0 | 25500 | 1.0870 | 0.8918 |
| 0.0 | 35.0 | 26250 | 1.0814 | 0.8935 |
| 0.0 | 36.0 | 27000 | 1.0837 | 0.8902 |
| 0.0038 | 37.0 | 27750 | 1.0867 | 0.8918 |
| 0.0 | 38.0 | 28500 | 1.1027 | 0.8918 |
| 0.0 | 39.0 | 29250 | 1.1059 | 0.8935 |
| 0.0 | 40.0 | 30000 | 1.1104 | 0.8918 |
| 0.0 | 41.0 | 30750 | 1.1167 | 0.8918 |
| 0.0 | 42.0 | 31500 | 1.1201 | 0.8918 |
| 0.0 | 43.0 | 32250 | 1.1220 | 0.8918 |
| 0.0 | 44.0 | 33000 | 1.1250 | 0.8918 |
| 0.0024 | 45.0 | 33750 | 1.1275 | 0.8918 |
| 0.0 | 46.0 | 34500 | 1.1293 | 0.8918 |
| 0.0 | 47.0 | 35250 | 1.1307 | 0.8918 |
| 0.0 | 48.0 | 36000 | 1.1319 | 0.8918 |
| 0.0 | 49.0 | 36750 | 1.1319 | 0.8918 |
| 0.0 | 50.0 | 37500 | 1.1317 | 0.8918 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2