smids_3x_deit_tiny_adamax_00001_fold3
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: 0.7809
- Accuracy: 0.9
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: 1e-05
- 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 |
|---|---|---|---|---|
| 0.4698 | 1.0 | 225 | 0.4392 | 0.8383 |
| 0.2874 | 2.0 | 450 | 0.3310 | 0.8683 |
| 0.2611 | 3.0 | 675 | 0.2984 | 0.8767 |
| 0.1948 | 4.0 | 900 | 0.2696 | 0.905 |
| 0.2186 | 5.0 | 1125 | 0.2803 | 0.905 |
| 0.1975 | 6.0 | 1350 | 0.2954 | 0.8983 |
| 0.1371 | 7.0 | 1575 | 0.2844 | 0.9033 |
| 0.1358 | 8.0 | 1800 | 0.2972 | 0.8983 |
| 0.1291 | 9.0 | 2025 | 0.3099 | 0.8967 |
| 0.0542 | 10.0 | 2250 | 0.3489 | 0.895 |
| 0.0501 | 11.0 | 2475 | 0.3613 | 0.8883 |
| 0.0606 | 12.0 | 2700 | 0.3786 | 0.8933 |
| 0.022 | 13.0 | 2925 | 0.4191 | 0.895 |
| 0.0131 | 14.0 | 3150 | 0.4369 | 0.8917 |
| 0.0269 | 15.0 | 3375 | 0.4976 | 0.89 |
| 0.017 | 16.0 | 3600 | 0.5139 | 0.8883 |
| 0.0239 | 17.0 | 3825 | 0.5627 | 0.905 |
| 0.0035 | 18.0 | 4050 | 0.5902 | 0.8933 |
| 0.0039 | 19.0 | 4275 | 0.6058 | 0.8967 |
| 0.0035 | 20.0 | 4500 | 0.6423 | 0.89 |
| 0.0003 | 21.0 | 4725 | 0.6358 | 0.8983 |
| 0.0002 | 22.0 | 4950 | 0.6169 | 0.9067 |
| 0.0002 | 23.0 | 5175 | 0.6520 | 0.8983 |
| 0.0001 | 24.0 | 5400 | 0.6716 | 0.8933 |
| 0.0214 | 25.0 | 5625 | 0.6822 | 0.8917 |
| 0.0001 | 26.0 | 5850 | 0.6829 | 0.895 |
| 0.0001 | 27.0 | 6075 | 0.7009 | 0.9017 |
| 0.0001 | 28.0 | 6300 | 0.7082 | 0.9033 |
| 0.0207 | 29.0 | 6525 | 0.7271 | 0.8967 |
| 0.0205 | 30.0 | 6750 | 0.7272 | 0.9033 |
| 0.0138 | 31.0 | 6975 | 0.7738 | 0.8867 |
| 0.0001 | 32.0 | 7200 | 0.7368 | 0.9033 |
| 0.0001 | 33.0 | 7425 | 0.7522 | 0.8967 |
| 0.0 | 34.0 | 7650 | 0.7497 | 0.8983 |
| 0.0 | 35.0 | 7875 | 0.7518 | 0.9017 |
| 0.0 | 36.0 | 8100 | 0.7530 | 0.9017 |
| 0.0 | 37.0 | 8325 | 0.7691 | 0.895 |
| 0.0 | 38.0 | 8550 | 0.7615 | 0.8983 |
| 0.0 | 39.0 | 8775 | 0.7639 | 0.9 |
| 0.0 | 40.0 | 9000 | 0.7671 | 0.9 |
| 0.0 | 41.0 | 9225 | 0.7823 | 0.8967 |
| 0.0 | 42.0 | 9450 | 0.7718 | 0.9 |
| 0.0 | 43.0 | 9675 | 0.7755 | 0.9 |
| 0.0 | 44.0 | 9900 | 0.7762 | 0.9017 |
| 0.0 | 45.0 | 10125 | 0.7825 | 0.8967 |
| 0.0 | 46.0 | 10350 | 0.7811 | 0.9 |
| 0.0044 | 47.0 | 10575 | 0.7795 | 0.9 |
| 0.0 | 48.0 | 10800 | 0.7809 | 0.9 |
| 0.0 | 49.0 | 11025 | 0.7812 | 0.9 |
| 0.0 | 50.0 | 11250 | 0.7809 | 0.9 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2