smids_3x_deit_small_adamax_001_fold1
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: 0.9284
- Accuracy: 0.8831
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: 0.001
- 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 |
|---|---|---|---|---|
| 0.5541 | 1.0 | 226 | 0.6212 | 0.7379 |
| 0.2991 | 2.0 | 452 | 0.3577 | 0.8748 |
| 0.3025 | 3.0 | 678 | 0.3328 | 0.8765 |
| 0.2535 | 4.0 | 904 | 0.3762 | 0.8631 |
| 0.2002 | 5.0 | 1130 | 0.3237 | 0.8581 |
| 0.2266 | 6.0 | 1356 | 0.2994 | 0.8898 |
| 0.1986 | 7.0 | 1582 | 0.4068 | 0.8831 |
| 0.102 | 8.0 | 1808 | 0.5862 | 0.8564 |
| 0.1677 | 9.0 | 2034 | 0.4842 | 0.8431 |
| 0.0849 | 10.0 | 2260 | 0.4657 | 0.8848 |
| 0.0597 | 11.0 | 2486 | 0.4867 | 0.8698 |
| 0.0739 | 12.0 | 2712 | 0.4601 | 0.8881 |
| 0.057 | 13.0 | 2938 | 0.5719 | 0.8614 |
| 0.0476 | 14.0 | 3164 | 0.5741 | 0.8898 |
| 0.0028 | 15.0 | 3390 | 0.5632 | 0.8932 |
| 0.0442 | 16.0 | 3616 | 0.6974 | 0.8748 |
| 0.0232 | 17.0 | 3842 | 0.7080 | 0.8731 |
| 0.0139 | 18.0 | 4068 | 0.6478 | 0.8848 |
| 0.0142 | 19.0 | 4294 | 0.6320 | 0.8881 |
| 0.0278 | 20.0 | 4520 | 0.7359 | 0.8698 |
| 0.0017 | 21.0 | 4746 | 0.5842 | 0.8948 |
| 0.0064 | 22.0 | 4972 | 0.8360 | 0.8681 |
| 0.0257 | 23.0 | 5198 | 0.6566 | 0.8865 |
| 0.0027 | 24.0 | 5424 | 0.6515 | 0.8948 |
| 0.0001 | 25.0 | 5650 | 0.6767 | 0.8998 |
| 0.0009 | 26.0 | 5876 | 0.7783 | 0.8865 |
| 0.0001 | 27.0 | 6102 | 0.7799 | 0.8815 |
| 0.0004 | 28.0 | 6328 | 0.8472 | 0.8831 |
| 0.0002 | 29.0 | 6554 | 0.6998 | 0.8848 |
| 0.0 | 30.0 | 6780 | 0.8253 | 0.8815 |
| 0.0 | 31.0 | 7006 | 0.7884 | 0.8898 |
| 0.0001 | 32.0 | 7232 | 0.8156 | 0.8798 |
| 0.0032 | 33.0 | 7458 | 0.8222 | 0.8881 |
| 0.0 | 34.0 | 7684 | 0.8187 | 0.8848 |
| 0.0 | 35.0 | 7910 | 0.8080 | 0.8848 |
| 0.0 | 36.0 | 8136 | 0.8285 | 0.8798 |
| 0.0 | 37.0 | 8362 | 0.8277 | 0.8848 |
| 0.0 | 38.0 | 8588 | 0.8358 | 0.8831 |
| 0.0 | 39.0 | 8814 | 0.8499 | 0.8831 |
| 0.0 | 40.0 | 9040 | 0.8554 | 0.8798 |
| 0.0034 | 41.0 | 9266 | 0.8762 | 0.8798 |
| 0.0035 | 42.0 | 9492 | 0.8827 | 0.8798 |
| 0.0 | 43.0 | 9718 | 0.8772 | 0.8798 |
| 0.0 | 44.0 | 9944 | 0.8964 | 0.8831 |
| 0.0 | 45.0 | 10170 | 0.9072 | 0.8831 |
| 0.0 | 46.0 | 10396 | 0.9164 | 0.8815 |
| 0.0 | 47.0 | 10622 | 0.9219 | 0.8815 |
| 0.0 | 48.0 | 10848 | 0.9247 | 0.8831 |
| 0.0 | 49.0 | 11074 | 0.9274 | 0.8831 |
| 0.0 | 50.0 | 11300 | 0.9284 | 0.8831 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2