smids_3x_deit_small_adamax_00001_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: 0.8130
- Accuracy: 0.8815
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 |
|---|---|---|---|---|
| 0.424 | 1.0 | 226 | 0.3513 | 0.8731 |
| 0.2946 | 2.0 | 452 | 0.3000 | 0.8798 |
| 0.1827 | 3.0 | 678 | 0.2979 | 0.8865 |
| 0.2004 | 4.0 | 904 | 0.3148 | 0.8932 |
| 0.1798 | 5.0 | 1130 | 0.3139 | 0.8998 |
| 0.0853 | 6.0 | 1356 | 0.3299 | 0.8965 |
| 0.0515 | 7.0 | 1582 | 0.3457 | 0.8865 |
| 0.0826 | 8.0 | 1808 | 0.3663 | 0.8932 |
| 0.0216 | 9.0 | 2034 | 0.3989 | 0.8948 |
| 0.0193 | 10.0 | 2260 | 0.4487 | 0.8948 |
| 0.0094 | 11.0 | 2486 | 0.5419 | 0.8815 |
| 0.0096 | 12.0 | 2712 | 0.5555 | 0.8898 |
| 0.0006 | 13.0 | 2938 | 0.5917 | 0.8881 |
| 0.0105 | 14.0 | 3164 | 0.6078 | 0.8982 |
| 0.0003 | 15.0 | 3390 | 0.6602 | 0.8815 |
| 0.0002 | 16.0 | 3616 | 0.6567 | 0.8815 |
| 0.0159 | 17.0 | 3842 | 0.6787 | 0.8881 |
| 0.0001 | 18.0 | 4068 | 0.7011 | 0.8865 |
| 0.0111 | 19.0 | 4294 | 0.7444 | 0.8815 |
| 0.0001 | 20.0 | 4520 | 0.7129 | 0.8865 |
| 0.0001 | 21.0 | 4746 | 0.7374 | 0.8898 |
| 0.0255 | 22.0 | 4972 | 0.7516 | 0.8748 |
| 0.0198 | 23.0 | 5198 | 0.7561 | 0.8898 |
| 0.0001 | 24.0 | 5424 | 0.7448 | 0.8865 |
| 0.0 | 25.0 | 5650 | 0.7702 | 0.8831 |
| 0.0001 | 26.0 | 5876 | 0.7830 | 0.8798 |
| 0.0 | 27.0 | 6102 | 0.7463 | 0.8898 |
| 0.0001 | 28.0 | 6328 | 0.7479 | 0.8815 |
| 0.0 | 29.0 | 6554 | 0.7622 | 0.8898 |
| 0.0 | 30.0 | 6780 | 0.7645 | 0.8898 |
| 0.0 | 31.0 | 7006 | 0.7711 | 0.8898 |
| 0.0 | 32.0 | 7232 | 0.7799 | 0.8865 |
| 0.012 | 33.0 | 7458 | 0.7796 | 0.8898 |
| 0.0 | 34.0 | 7684 | 0.8089 | 0.8848 |
| 0.0 | 35.0 | 7910 | 0.7888 | 0.8831 |
| 0.0 | 36.0 | 8136 | 0.8152 | 0.8815 |
| 0.0 | 37.0 | 8362 | 0.7978 | 0.8831 |
| 0.0 | 38.0 | 8588 | 0.7987 | 0.8781 |
| 0.0 | 39.0 | 8814 | 0.7888 | 0.8915 |
| 0.0 | 40.0 | 9040 | 0.7970 | 0.8831 |
| 0.0095 | 41.0 | 9266 | 0.8023 | 0.8815 |
| 0.0008 | 42.0 | 9492 | 0.8065 | 0.8798 |
| 0.0 | 43.0 | 9718 | 0.7964 | 0.8881 |
| 0.0 | 44.0 | 9944 | 0.8002 | 0.8865 |
| 0.0 | 45.0 | 10170 | 0.8035 | 0.8831 |
| 0.0 | 46.0 | 10396 | 0.8056 | 0.8815 |
| 0.0 | 47.0 | 10622 | 0.8110 | 0.8815 |
| 0.0 | 48.0 | 10848 | 0.8125 | 0.8831 |
| 0.0 | 49.0 | 11074 | 0.8131 | 0.8815 |
| 0.0 | 50.0 | 11300 | 0.8130 | 0.8815 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2