smids_10x_deit_small_adamax_00001_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: 0.8990
- Accuracy: 0.8967
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 |
|---|---|---|---|---|
| 0.2314 | 1.0 | 750 | 0.3103 | 0.8783 |
| 0.2018 | 2.0 | 1500 | 0.2701 | 0.8983 |
| 0.1509 | 3.0 | 2250 | 0.2783 | 0.9 |
| 0.0994 | 4.0 | 3000 | 0.3003 | 0.8983 |
| 0.1044 | 5.0 | 3750 | 0.3361 | 0.905 |
| 0.0438 | 6.0 | 4500 | 0.4330 | 0.89 |
| 0.017 | 7.0 | 5250 | 0.4929 | 0.9 |
| 0.0271 | 8.0 | 6000 | 0.5583 | 0.8883 |
| 0.0025 | 9.0 | 6750 | 0.6641 | 0.895 |
| 0.0019 | 10.0 | 7500 | 0.6642 | 0.895 |
| 0.0069 | 11.0 | 8250 | 0.7364 | 0.895 |
| 0.0001 | 12.0 | 9000 | 0.7531 | 0.8917 |
| 0.0001 | 13.0 | 9750 | 0.7730 | 0.8983 |
| 0.0 | 14.0 | 10500 | 0.7588 | 0.895 |
| 0.0 | 15.0 | 11250 | 0.8009 | 0.8933 |
| 0.0 | 16.0 | 12000 | 0.8380 | 0.8883 |
| 0.002 | 17.0 | 12750 | 0.8384 | 0.8917 |
| 0.0 | 18.0 | 13500 | 0.8173 | 0.9017 |
| 0.0 | 19.0 | 14250 | 0.8127 | 0.8983 |
| 0.0 | 20.0 | 15000 | 0.8957 | 0.8917 |
| 0.0 | 21.0 | 15750 | 0.8587 | 0.8917 |
| 0.0 | 22.0 | 16500 | 0.8745 | 0.89 |
| 0.0 | 23.0 | 17250 | 0.8882 | 0.8917 |
| 0.0 | 24.0 | 18000 | 0.8669 | 0.895 |
| 0.0 | 25.0 | 18750 | 0.8973 | 0.8917 |
| 0.0 | 26.0 | 19500 | 0.8589 | 0.895 |
| 0.0 | 27.0 | 20250 | 0.8360 | 0.9 |
| 0.0 | 28.0 | 21000 | 0.8787 | 0.8967 |
| 0.0 | 29.0 | 21750 | 0.8640 | 0.895 |
| 0.0039 | 30.0 | 22500 | 0.8923 | 0.8867 |
| 0.0 | 31.0 | 23250 | 0.8650 | 0.8967 |
| 0.0 | 32.0 | 24000 | 0.8868 | 0.8983 |
| 0.0 | 33.0 | 24750 | 0.8702 | 0.8967 |
| 0.0 | 34.0 | 25500 | 0.8705 | 0.8933 |
| 0.0 | 35.0 | 26250 | 0.8693 | 0.8933 |
| 0.0 | 36.0 | 27000 | 0.8764 | 0.895 |
| 0.0035 | 37.0 | 27750 | 0.8971 | 0.8933 |
| 0.0 | 38.0 | 28500 | 0.8853 | 0.895 |
| 0.0 | 39.0 | 29250 | 0.8831 | 0.895 |
| 0.0 | 40.0 | 30000 | 0.8852 | 0.895 |
| 0.0 | 41.0 | 30750 | 0.8936 | 0.8967 |
| 0.0 | 42.0 | 31500 | 0.8989 | 0.895 |
| 0.0 | 43.0 | 32250 | 0.8952 | 0.895 |
| 0.0 | 44.0 | 33000 | 0.8982 | 0.895 |
| 0.0 | 45.0 | 33750 | 0.8981 | 0.895 |
| 0.0 | 46.0 | 34500 | 0.8978 | 0.895 |
| 0.0 | 47.0 | 35250 | 0.8987 | 0.895 |
| 0.0 | 48.0 | 36000 | 0.8989 | 0.895 |
| 0.0 | 49.0 | 36750 | 0.8995 | 0.895 |
| 0.0 | 50.0 | 37500 | 0.8990 | 0.8967 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2