smids_10x_deit_tiny_sgd_001_fold4
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: 0.3938
- Accuracy: 0.8667
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 |
|---|---|---|---|---|
| 0.5453 | 1.0 | 750 | 0.5623 | 0.7633 |
| 0.3882 | 2.0 | 1500 | 0.4483 | 0.8183 |
| 0.3799 | 3.0 | 2250 | 0.4088 | 0.8317 |
| 0.3643 | 4.0 | 3000 | 0.3893 | 0.8383 |
| 0.2628 | 5.0 | 3750 | 0.3770 | 0.8467 |
| 0.2344 | 6.0 | 4500 | 0.3757 | 0.8467 |
| 0.2158 | 7.0 | 5250 | 0.3640 | 0.8583 |
| 0.2518 | 8.0 | 6000 | 0.3700 | 0.86 |
| 0.2784 | 9.0 | 6750 | 0.3645 | 0.8617 |
| 0.2124 | 10.0 | 7500 | 0.3619 | 0.86 |
| 0.2508 | 11.0 | 8250 | 0.3628 | 0.8583 |
| 0.2963 | 12.0 | 9000 | 0.3717 | 0.86 |
| 0.2464 | 13.0 | 9750 | 0.3675 | 0.86 |
| 0.2153 | 14.0 | 10500 | 0.3661 | 0.8633 |
| 0.1783 | 15.0 | 11250 | 0.3637 | 0.8633 |
| 0.1889 | 16.0 | 12000 | 0.3675 | 0.865 |
| 0.1615 | 17.0 | 12750 | 0.3615 | 0.8633 |
| 0.1602 | 18.0 | 13500 | 0.3665 | 0.8683 |
| 0.2382 | 19.0 | 14250 | 0.3640 | 0.8633 |
| 0.1431 | 20.0 | 15000 | 0.3640 | 0.8667 |
| 0.1246 | 21.0 | 15750 | 0.3698 | 0.865 |
| 0.1642 | 22.0 | 16500 | 0.3698 | 0.8617 |
| 0.1435 | 23.0 | 17250 | 0.3719 | 0.8617 |
| 0.184 | 24.0 | 18000 | 0.3745 | 0.865 |
| 0.1543 | 25.0 | 18750 | 0.3749 | 0.8617 |
| 0.1463 | 26.0 | 19500 | 0.3762 | 0.8633 |
| 0.1225 | 27.0 | 20250 | 0.3737 | 0.8667 |
| 0.1542 | 28.0 | 21000 | 0.3785 | 0.865 |
| 0.1065 | 29.0 | 21750 | 0.3788 | 0.87 |
| 0.1351 | 30.0 | 22500 | 0.3799 | 0.8667 |
| 0.1281 | 31.0 | 23250 | 0.3825 | 0.8667 |
| 0.1337 | 32.0 | 24000 | 0.3866 | 0.8633 |
| 0.1066 | 33.0 | 24750 | 0.3848 | 0.8667 |
| 0.1503 | 34.0 | 25500 | 0.3856 | 0.87 |
| 0.0933 | 35.0 | 26250 | 0.3837 | 0.8717 |
| 0.1119 | 36.0 | 27000 | 0.3871 | 0.87 |
| 0.0916 | 37.0 | 27750 | 0.3845 | 0.87 |
| 0.1419 | 38.0 | 28500 | 0.3888 | 0.8683 |
| 0.1831 | 39.0 | 29250 | 0.3865 | 0.87 |
| 0.1443 | 40.0 | 30000 | 0.3886 | 0.8683 |
| 0.1089 | 41.0 | 30750 | 0.3938 | 0.865 |
| 0.0931 | 42.0 | 31500 | 0.3903 | 0.8683 |
| 0.1349 | 43.0 | 32250 | 0.3917 | 0.8683 |
| 0.1005 | 44.0 | 33000 | 0.3917 | 0.8667 |
| 0.12 | 45.0 | 33750 | 0.3918 | 0.8667 |
| 0.1354 | 46.0 | 34500 | 0.3924 | 0.8667 |
| 0.0817 | 47.0 | 35250 | 0.3922 | 0.8667 |
| 0.0828 | 48.0 | 36000 | 0.3931 | 0.8667 |
| 0.0941 | 49.0 | 36750 | 0.3938 | 0.8667 |
| 0.0837 | 50.0 | 37500 | 0.3938 | 0.8667 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2