smids_10x_deit_small_rms_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: 1.0878
- Accuracy: 0.8968
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 |
|---|---|---|---|---|
| 0.2199 | 1.0 | 750 | 0.3087 | 0.8702 |
| 0.1415 | 2.0 | 1500 | 0.3270 | 0.8952 |
| 0.0884 | 3.0 | 2250 | 0.4634 | 0.8735 |
| 0.1125 | 4.0 | 3000 | 0.5598 | 0.8735 |
| 0.0577 | 5.0 | 3750 | 0.5910 | 0.8785 |
| 0.0338 | 6.0 | 4500 | 0.6886 | 0.8802 |
| 0.0535 | 7.0 | 5250 | 0.5872 | 0.8835 |
| 0.0197 | 8.0 | 6000 | 0.6155 | 0.8952 |
| 0.0062 | 9.0 | 6750 | 0.7198 | 0.8852 |
| 0.019 | 10.0 | 7500 | 0.6897 | 0.8985 |
| 0.011 | 11.0 | 8250 | 0.6348 | 0.8985 |
| 0.0457 | 12.0 | 9000 | 0.6894 | 0.8835 |
| 0.0048 | 13.0 | 9750 | 0.7305 | 0.8918 |
| 0.0312 | 14.0 | 10500 | 0.8235 | 0.8835 |
| 0.0002 | 15.0 | 11250 | 0.8886 | 0.8819 |
| 0.0307 | 16.0 | 12000 | 0.8040 | 0.8769 |
| 0.0004 | 17.0 | 12750 | 0.6236 | 0.9068 |
| 0.0012 | 18.0 | 13500 | 0.8211 | 0.8785 |
| 0.0004 | 19.0 | 14250 | 0.6599 | 0.8985 |
| 0.0198 | 20.0 | 15000 | 0.6886 | 0.8985 |
| 0.0006 | 21.0 | 15750 | 0.8044 | 0.8852 |
| 0.007 | 22.0 | 16500 | 0.7019 | 0.8885 |
| 0.0333 | 23.0 | 17250 | 0.7287 | 0.8819 |
| 0.0 | 24.0 | 18000 | 1.0716 | 0.8652 |
| 0.0 | 25.0 | 18750 | 0.9627 | 0.8752 |
| 0.0006 | 26.0 | 19500 | 1.0237 | 0.8686 |
| 0.0027 | 27.0 | 20250 | 0.9748 | 0.8769 |
| 0.0004 | 28.0 | 21000 | 0.9776 | 0.8902 |
| 0.0 | 29.0 | 21750 | 0.9254 | 0.8785 |
| 0.0001 | 30.0 | 22500 | 0.9772 | 0.8902 |
| 0.0132 | 31.0 | 23250 | 0.7890 | 0.9035 |
| 0.0002 | 32.0 | 24000 | 0.8329 | 0.8985 |
| 0.0 | 33.0 | 24750 | 0.8259 | 0.9101 |
| 0.0 | 34.0 | 25500 | 0.9870 | 0.8852 |
| 0.0067 | 35.0 | 26250 | 1.0178 | 0.8918 |
| 0.0 | 36.0 | 27000 | 0.9706 | 0.9035 |
| 0.0039 | 37.0 | 27750 | 0.9405 | 0.8952 |
| 0.0 | 38.0 | 28500 | 1.0909 | 0.8869 |
| 0.0 | 39.0 | 29250 | 1.0161 | 0.9002 |
| 0.0 | 40.0 | 30000 | 0.9672 | 0.9068 |
| 0.0 | 41.0 | 30750 | 1.0373 | 0.8952 |
| 0.0 | 42.0 | 31500 | 1.0954 | 0.8968 |
| 0.0 | 43.0 | 32250 | 1.0521 | 0.9035 |
| 0.0 | 44.0 | 33000 | 1.0547 | 0.8968 |
| 0.0028 | 45.0 | 33750 | 1.0695 | 0.8952 |
| 0.0 | 46.0 | 34500 | 1.0676 | 0.8968 |
| 0.0 | 47.0 | 35250 | 1.0805 | 0.8968 |
| 0.0 | 48.0 | 36000 | 1.0830 | 0.8968 |
| 0.0 | 49.0 | 36750 | 1.0856 | 0.8968 |
| 0.0 | 50.0 | 37500 | 1.0878 | 0.8968 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2