smids_3x_deit_tiny_sgd_001_fold2
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: 0.3631
- Accuracy: 0.8652
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 |
|---|---|---|---|---|
| 0.9177 | 1.0 | 225 | 0.8996 | 0.5691 |
| 0.6997 | 2.0 | 450 | 0.6912 | 0.7121 |
| 0.5229 | 3.0 | 675 | 0.5718 | 0.7671 |
| 0.5533 | 4.0 | 900 | 0.5111 | 0.8020 |
| 0.4272 | 5.0 | 1125 | 0.4697 | 0.8070 |
| 0.3877 | 6.0 | 1350 | 0.4425 | 0.8170 |
| 0.4004 | 7.0 | 1575 | 0.4203 | 0.8336 |
| 0.3661 | 8.0 | 1800 | 0.4043 | 0.8369 |
| 0.3402 | 9.0 | 2025 | 0.3983 | 0.8386 |
| 0.2899 | 10.0 | 2250 | 0.3839 | 0.8486 |
| 0.3594 | 11.0 | 2475 | 0.3760 | 0.8469 |
| 0.2789 | 12.0 | 2700 | 0.3717 | 0.8502 |
| 0.2808 | 13.0 | 2925 | 0.3681 | 0.8502 |
| 0.2912 | 14.0 | 3150 | 0.3664 | 0.8552 |
| 0.2944 | 15.0 | 3375 | 0.3661 | 0.8502 |
| 0.3273 | 16.0 | 3600 | 0.3677 | 0.8552 |
| 0.2474 | 17.0 | 3825 | 0.3614 | 0.8552 |
| 0.1928 | 18.0 | 4050 | 0.3628 | 0.8569 |
| 0.2096 | 19.0 | 4275 | 0.3553 | 0.8519 |
| 0.2614 | 20.0 | 4500 | 0.3573 | 0.8552 |
| 0.2898 | 21.0 | 4725 | 0.3557 | 0.8619 |
| 0.3219 | 22.0 | 4950 | 0.3582 | 0.8536 |
| 0.3025 | 23.0 | 5175 | 0.3562 | 0.8602 |
| 0.28 | 24.0 | 5400 | 0.3553 | 0.8569 |
| 0.2538 | 25.0 | 5625 | 0.3547 | 0.8569 |
| 0.2485 | 26.0 | 5850 | 0.3551 | 0.8586 |
| 0.2246 | 27.0 | 6075 | 0.3556 | 0.8619 |
| 0.2303 | 28.0 | 6300 | 0.3556 | 0.8602 |
| 0.2272 | 29.0 | 6525 | 0.3568 | 0.8619 |
| 0.2494 | 30.0 | 6750 | 0.3572 | 0.8602 |
| 0.1942 | 31.0 | 6975 | 0.3593 | 0.8619 |
| 0.2095 | 32.0 | 7200 | 0.3591 | 0.8619 |
| 0.2432 | 33.0 | 7425 | 0.3587 | 0.8619 |
| 0.2713 | 34.0 | 7650 | 0.3578 | 0.8586 |
| 0.1998 | 35.0 | 7875 | 0.3599 | 0.8619 |
| 0.2229 | 36.0 | 8100 | 0.3607 | 0.8586 |
| 0.2109 | 37.0 | 8325 | 0.3599 | 0.8619 |
| 0.1909 | 38.0 | 8550 | 0.3609 | 0.8602 |
| 0.1902 | 39.0 | 8775 | 0.3619 | 0.8586 |
| 0.2221 | 40.0 | 9000 | 0.3623 | 0.8586 |
| 0.1747 | 41.0 | 9225 | 0.3610 | 0.8586 |
| 0.1796 | 42.0 | 9450 | 0.3605 | 0.8602 |
| 0.1695 | 43.0 | 9675 | 0.3624 | 0.8619 |
| 0.2018 | 44.0 | 9900 | 0.3615 | 0.8619 |
| 0.2591 | 45.0 | 10125 | 0.3627 | 0.8602 |
| 0.2 | 46.0 | 10350 | 0.3630 | 0.8602 |
| 0.1903 | 47.0 | 10575 | 0.3635 | 0.8619 |
| 0.1709 | 48.0 | 10800 | 0.3630 | 0.8636 |
| 0.21 | 49.0 | 11025 | 0.3631 | 0.8636 |
| 0.168 | 50.0 | 11250 | 0.3631 | 0.8652 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2