smids_3x_deit_small_adamax_0001_fold5
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: 0.8808
- 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: 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 |
|---|---|---|---|---|
| 0.3174 | 1.0 | 225 | 0.2998 | 0.8783 |
| 0.1192 | 2.0 | 450 | 0.2723 | 0.9083 |
| 0.0978 | 3.0 | 675 | 0.4594 | 0.8617 |
| 0.0329 | 4.0 | 900 | 0.4368 | 0.8833 |
| 0.0356 | 5.0 | 1125 | 0.4938 | 0.89 |
| 0.0115 | 6.0 | 1350 | 0.5783 | 0.895 |
| 0.0594 | 7.0 | 1575 | 0.5882 | 0.8933 |
| 0.0073 | 8.0 | 1800 | 0.6931 | 0.8983 |
| 0.0161 | 9.0 | 2025 | 0.7537 | 0.8917 |
| 0.0716 | 10.0 | 2250 | 0.8442 | 0.885 |
| 0.0028 | 11.0 | 2475 | 0.9224 | 0.875 |
| 0.0418 | 12.0 | 2700 | 0.8061 | 0.8883 |
| 0.0025 | 13.0 | 2925 | 0.9720 | 0.865 |
| 0.0001 | 14.0 | 3150 | 0.8224 | 0.8883 |
| 0.0003 | 15.0 | 3375 | 0.8506 | 0.8883 |
| 0.0 | 16.0 | 3600 | 0.8357 | 0.8883 |
| 0.0 | 17.0 | 3825 | 0.7488 | 0.895 |
| 0.0043 | 18.0 | 4050 | 0.8378 | 0.8783 |
| 0.008 | 19.0 | 4275 | 0.8339 | 0.8883 |
| 0.0009 | 20.0 | 4500 | 0.8302 | 0.8917 |
| 0.0 | 21.0 | 4725 | 0.8494 | 0.89 |
| 0.0 | 22.0 | 4950 | 0.8420 | 0.895 |
| 0.0 | 23.0 | 5175 | 0.8388 | 0.8933 |
| 0.0042 | 24.0 | 5400 | 0.8556 | 0.895 |
| 0.0 | 25.0 | 5625 | 0.8527 | 0.895 |
| 0.0 | 26.0 | 5850 | 0.8812 | 0.8917 |
| 0.0 | 27.0 | 6075 | 0.8957 | 0.8917 |
| 0.0 | 28.0 | 6300 | 0.8419 | 0.8933 |
| 0.0034 | 29.0 | 6525 | 0.8176 | 0.8967 |
| 0.0 | 30.0 | 6750 | 0.8204 | 0.9 |
| 0.0 | 31.0 | 6975 | 0.8207 | 0.9 |
| 0.0 | 32.0 | 7200 | 0.8298 | 0.8933 |
| 0.0034 | 33.0 | 7425 | 0.8331 | 0.8967 |
| 0.0 | 34.0 | 7650 | 0.8399 | 0.895 |
| 0.0029 | 35.0 | 7875 | 0.8427 | 0.895 |
| 0.0 | 36.0 | 8100 | 0.8464 | 0.895 |
| 0.004 | 37.0 | 8325 | 0.8513 | 0.8967 |
| 0.0 | 38.0 | 8550 | 0.8556 | 0.8983 |
| 0.0 | 39.0 | 8775 | 0.8574 | 0.8933 |
| 0.0 | 40.0 | 9000 | 0.8619 | 0.8967 |
| 0.0 | 41.0 | 9225 | 0.8650 | 0.8983 |
| 0.0 | 42.0 | 9450 | 0.8684 | 0.8983 |
| 0.0027 | 43.0 | 9675 | 0.8709 | 0.8983 |
| 0.0 | 44.0 | 9900 | 0.8740 | 0.8983 |
| 0.0 | 45.0 | 10125 | 0.8773 | 0.8967 |
| 0.0 | 46.0 | 10350 | 0.8777 | 0.9 |
| 0.0 | 47.0 | 10575 | 0.8792 | 0.9 |
| 0.0 | 48.0 | 10800 | 0.8800 | 0.9 |
| 0.0 | 49.0 | 11025 | 0.8807 | 0.9 |
| 0.0 | 50.0 | 11250 | 0.8808 | 0.9 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2