smids_5x_deit_small_adamax_0001_fold3
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: 1.0059
- Accuracy: 0.8983
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 |
|---|---|---|---|---|
| 0.1525 | 1.0 | 750 | 0.2697 | 0.895 |
| 0.1282 | 2.0 | 1500 | 0.3452 | 0.8917 |
| 0.0382 | 3.0 | 2250 | 0.4781 | 0.89 |
| 0.0664 | 4.0 | 3000 | 0.6005 | 0.8917 |
| 0.0139 | 5.0 | 3750 | 0.6269 | 0.8983 |
| 0.0003 | 6.0 | 4500 | 0.7710 | 0.9017 |
| 0.0029 | 7.0 | 5250 | 0.7996 | 0.89 |
| 0.0177 | 8.0 | 6000 | 0.7898 | 0.9033 |
| 0.0001 | 9.0 | 6750 | 0.7240 | 0.8983 |
| 0.0001 | 10.0 | 7500 | 0.8037 | 0.9133 |
| 0.0 | 11.0 | 8250 | 0.7846 | 0.91 |
| 0.0005 | 12.0 | 9000 | 0.9174 | 0.885 |
| 0.04 | 13.0 | 9750 | 0.8629 | 0.8983 |
| 0.0005 | 14.0 | 10500 | 0.8319 | 0.895 |
| 0.0 | 15.0 | 11250 | 0.8174 | 0.91 |
| 0.0 | 16.0 | 12000 | 0.8650 | 0.9017 |
| 0.0 | 17.0 | 12750 | 0.7601 | 0.9133 |
| 0.0 | 18.0 | 13500 | 0.9296 | 0.8867 |
| 0.0 | 19.0 | 14250 | 0.8663 | 0.9033 |
| 0.0 | 20.0 | 15000 | 0.9126 | 0.895 |
| 0.0 | 21.0 | 15750 | 0.8974 | 0.8983 |
| 0.0 | 22.0 | 16500 | 0.9203 | 0.8967 |
| 0.0 | 23.0 | 17250 | 0.8786 | 0.9033 |
| 0.0 | 24.0 | 18000 | 0.8565 | 0.9017 |
| 0.0 | 25.0 | 18750 | 0.9193 | 0.89 |
| 0.0 | 26.0 | 19500 | 0.9069 | 0.895 |
| 0.0 | 27.0 | 20250 | 0.8841 | 0.9 |
| 0.0 | 28.0 | 21000 | 0.9282 | 0.895 |
| 0.0 | 29.0 | 21750 | 0.9329 | 0.8983 |
| 0.0 | 30.0 | 22500 | 0.9485 | 0.9017 |
| 0.0 | 31.0 | 23250 | 0.9410 | 0.8967 |
| 0.0 | 32.0 | 24000 | 0.9299 | 0.9 |
| 0.0 | 33.0 | 24750 | 0.9416 | 0.8983 |
| 0.0 | 34.0 | 25500 | 0.9468 | 0.8967 |
| 0.0 | 35.0 | 26250 | 0.9697 | 0.895 |
| 0.0 | 36.0 | 27000 | 0.9684 | 0.8983 |
| 0.0 | 37.0 | 27750 | 0.9718 | 0.8983 |
| 0.0 | 38.0 | 28500 | 0.9758 | 0.8983 |
| 0.0 | 39.0 | 29250 | 0.9793 | 0.8983 |
| 0.0 | 40.0 | 30000 | 0.9881 | 0.8983 |
| 0.0 | 41.0 | 30750 | 0.9875 | 0.8983 |
| 0.0 | 42.0 | 31500 | 0.9984 | 0.8983 |
| 0.0 | 43.0 | 32250 | 0.9995 | 0.8983 |
| 0.0 | 44.0 | 33000 | 1.0002 | 0.8983 |
| 0.0 | 45.0 | 33750 | 1.0011 | 0.8983 |
| 0.0 | 46.0 | 34500 | 1.0026 | 0.8983 |
| 0.0 | 47.0 | 35250 | 1.0030 | 0.8983 |
| 0.0 | 48.0 | 36000 | 1.0034 | 0.8983 |
| 0.0 | 49.0 | 36750 | 1.0045 | 0.8983 |
| 0.0 | 50.0 | 37500 | 1.0059 | 0.8983 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2