smids_5x_deit_tiny_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: 0.3358
- Accuracy: 0.8686
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 |
|---|---|---|---|---|
| 0.7754 | 1.0 | 375 | 0.7240 | 0.7271 |
| 0.5298 | 2.0 | 750 | 0.5482 | 0.7837 |
| 0.4453 | 3.0 | 1125 | 0.4761 | 0.8186 |
| 0.3233 | 4.0 | 1500 | 0.4354 | 0.8286 |
| 0.3301 | 5.0 | 1875 | 0.4115 | 0.8386 |
| 0.3179 | 6.0 | 2250 | 0.3924 | 0.8469 |
| 0.3101 | 7.0 | 2625 | 0.3803 | 0.8502 |
| 0.3266 | 8.0 | 3000 | 0.3685 | 0.8586 |
| 0.2663 | 9.0 | 3375 | 0.3605 | 0.8552 |
| 0.2805 | 10.0 | 3750 | 0.3550 | 0.8536 |
| 0.2677 | 11.0 | 4125 | 0.3495 | 0.8619 |
| 0.3046 | 12.0 | 4500 | 0.3461 | 0.8686 |
| 0.2173 | 13.0 | 4875 | 0.3409 | 0.8602 |
| 0.2384 | 14.0 | 5250 | 0.3398 | 0.8636 |
| 0.2681 | 15.0 | 5625 | 0.3343 | 0.8652 |
| 0.1901 | 16.0 | 6000 | 0.3336 | 0.8735 |
| 0.2623 | 17.0 | 6375 | 0.3353 | 0.8735 |
| 0.1865 | 18.0 | 6750 | 0.3314 | 0.8735 |
| 0.2003 | 19.0 | 7125 | 0.3309 | 0.8735 |
| 0.2713 | 20.0 | 7500 | 0.3280 | 0.8752 |
| 0.2017 | 21.0 | 7875 | 0.3298 | 0.8702 |
| 0.1863 | 22.0 | 8250 | 0.3281 | 0.8769 |
| 0.227 | 23.0 | 8625 | 0.3271 | 0.8769 |
| 0.1889 | 24.0 | 9000 | 0.3290 | 0.8752 |
| 0.1561 | 25.0 | 9375 | 0.3282 | 0.8752 |
| 0.2339 | 26.0 | 9750 | 0.3258 | 0.8752 |
| 0.2006 | 27.0 | 10125 | 0.3286 | 0.8802 |
| 0.1745 | 28.0 | 10500 | 0.3294 | 0.8719 |
| 0.1852 | 29.0 | 10875 | 0.3284 | 0.8719 |
| 0.1931 | 30.0 | 11250 | 0.3301 | 0.8702 |
| 0.1811 | 31.0 | 11625 | 0.3297 | 0.8735 |
| 0.1783 | 32.0 | 12000 | 0.3325 | 0.8702 |
| 0.1809 | 33.0 | 12375 | 0.3288 | 0.8769 |
| 0.1274 | 34.0 | 12750 | 0.3315 | 0.8652 |
| 0.1957 | 35.0 | 13125 | 0.3314 | 0.8702 |
| 0.1704 | 36.0 | 13500 | 0.3319 | 0.8686 |
| 0.1796 | 37.0 | 13875 | 0.3309 | 0.8686 |
| 0.1565 | 38.0 | 14250 | 0.3327 | 0.8702 |
| 0.1735 | 39.0 | 14625 | 0.3325 | 0.8686 |
| 0.1525 | 40.0 | 15000 | 0.3345 | 0.8669 |
| 0.1548 | 41.0 | 15375 | 0.3344 | 0.8735 |
| 0.1677 | 42.0 | 15750 | 0.3353 | 0.8669 |
| 0.1708 | 43.0 | 16125 | 0.3357 | 0.8669 |
| 0.1467 | 44.0 | 16500 | 0.3356 | 0.8669 |
| 0.1338 | 45.0 | 16875 | 0.3358 | 0.8686 |
| 0.2032 | 46.0 | 17250 | 0.3360 | 0.8669 |
| 0.1609 | 47.0 | 17625 | 0.3359 | 0.8686 |
| 0.155 | 48.0 | 18000 | 0.3359 | 0.8686 |
| 0.2258 | 49.0 | 18375 | 0.3359 | 0.8669 |
| 0.1319 | 50.0 | 18750 | 0.3358 | 0.8686 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2