smids_10x_deit_tiny_rms_00001_fold1
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: 0.9553
- Accuracy: 0.9115
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 |
|---|---|---|---|---|
| 0.2396 | 1.0 | 751 | 0.2956 | 0.8932 |
| 0.1563 | 2.0 | 1502 | 0.3511 | 0.8748 |
| 0.1378 | 3.0 | 2253 | 0.3209 | 0.9032 |
| 0.0749 | 4.0 | 3004 | 0.4152 | 0.8965 |
| 0.0902 | 5.0 | 3755 | 0.4990 | 0.8998 |
| 0.0311 | 6.0 | 4506 | 0.6972 | 0.8948 |
| 0.0511 | 7.0 | 5257 | 0.6707 | 0.8915 |
| 0.0692 | 8.0 | 6008 | 0.7791 | 0.8948 |
| 0.0655 | 9.0 | 6759 | 0.7801 | 0.8965 |
| 0.0191 | 10.0 | 7510 | 0.8995 | 0.8948 |
| 0.0314 | 11.0 | 8261 | 0.8069 | 0.8965 |
| 0.0092 | 12.0 | 9012 | 0.8789 | 0.8881 |
| 0.0207 | 13.0 | 9763 | 0.9080 | 0.8915 |
| 0.0001 | 14.0 | 10514 | 0.9075 | 0.8982 |
| 0.0137 | 15.0 | 11265 | 1.1216 | 0.8915 |
| 0.0289 | 16.0 | 12016 | 1.0268 | 0.8932 |
| 0.0749 | 17.0 | 12767 | 0.9684 | 0.8965 |
| 0.0004 | 18.0 | 13518 | 0.9374 | 0.8948 |
| 0.0016 | 19.0 | 14269 | 0.9146 | 0.8998 |
| 0.0 | 20.0 | 15020 | 0.8660 | 0.9115 |
| 0.0 | 21.0 | 15771 | 0.8768 | 0.9132 |
| 0.0205 | 22.0 | 16522 | 0.9737 | 0.8932 |
| 0.0 | 23.0 | 17273 | 0.8857 | 0.9065 |
| 0.0 | 24.0 | 18024 | 0.9206 | 0.9015 |
| 0.0 | 25.0 | 18775 | 0.9882 | 0.9032 |
| 0.0051 | 26.0 | 19526 | 0.9311 | 0.9048 |
| 0.0005 | 27.0 | 20277 | 0.9916 | 0.8831 |
| 0.0 | 28.0 | 21028 | 0.8978 | 0.9065 |
| 0.0147 | 29.0 | 21779 | 0.9817 | 0.8998 |
| 0.0 | 30.0 | 22530 | 0.9072 | 0.9132 |
| 0.0 | 31.0 | 23281 | 0.9000 | 0.9032 |
| 0.0 | 32.0 | 24032 | 0.9908 | 0.9048 |
| 0.0 | 33.0 | 24783 | 0.9477 | 0.8998 |
| 0.0001 | 34.0 | 25534 | 0.9361 | 0.8998 |
| 0.0 | 35.0 | 26285 | 0.9285 | 0.9048 |
| 0.0 | 36.0 | 27036 | 0.9622 | 0.9048 |
| 0.0 | 37.0 | 27787 | 0.9080 | 0.9082 |
| 0.0 | 38.0 | 28538 | 1.0318 | 0.9065 |
| 0.0 | 39.0 | 29289 | 0.8954 | 0.9115 |
| 0.0 | 40.0 | 30040 | 0.9047 | 0.9098 |
| 0.0 | 41.0 | 30791 | 0.9568 | 0.9098 |
| 0.0 | 42.0 | 31542 | 0.9648 | 0.9082 |
| 0.0 | 43.0 | 32293 | 0.9575 | 0.9098 |
| 0.0 | 44.0 | 33044 | 0.9498 | 0.9132 |
| 0.0 | 45.0 | 33795 | 0.9583 | 0.9098 |
| 0.0 | 46.0 | 34546 | 0.9523 | 0.9115 |
| 0.0 | 47.0 | 35297 | 0.9525 | 0.9115 |
| 0.0 | 48.0 | 36048 | 0.9545 | 0.9115 |
| 0.0 | 49.0 | 36799 | 0.9543 | 0.9115 |
| 0.0 | 50.0 | 37550 | 0.9553 | 0.9115 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2