hushem_5x_deit_tiny_rms_0001_fold2
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: 1.6984
- Accuracy: 0.8222
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 |
|---|---|---|---|---|
| 1.4321 | 1.0 | 27 | 1.3812 | 0.2444 |
| 1.2336 | 2.0 | 54 | 1.7089 | 0.3556 |
| 0.6405 | 3.0 | 81 | 1.3102 | 0.5333 |
| 0.4623 | 4.0 | 108 | 1.3984 | 0.5778 |
| 0.2446 | 5.0 | 135 | 1.1033 | 0.7111 |
| 0.1889 | 6.0 | 162 | 1.1501 | 0.7778 |
| 0.1099 | 7.0 | 189 | 1.0953 | 0.7778 |
| 0.0445 | 8.0 | 216 | 0.8781 | 0.8444 |
| 0.0952 | 9.0 | 243 | 1.4626 | 0.8222 |
| 0.0463 | 10.0 | 270 | 1.4665 | 0.7778 |
| 0.0126 | 11.0 | 297 | 1.5722 | 0.7556 |
| 0.0011 | 12.0 | 324 | 1.5320 | 0.7778 |
| 0.0001 | 13.0 | 351 | 1.4353 | 0.7778 |
| 0.0001 | 14.0 | 378 | 1.4432 | 0.8 |
| 0.0001 | 15.0 | 405 | 1.4496 | 0.8 |
| 0.0 | 16.0 | 432 | 1.4554 | 0.8 |
| 0.0 | 17.0 | 459 | 1.4609 | 0.8 |
| 0.0 | 18.0 | 486 | 1.4660 | 0.8 |
| 0.0 | 19.0 | 513 | 1.4713 | 0.8 |
| 0.0 | 20.0 | 540 | 1.4786 | 0.8 |
| 0.0 | 21.0 | 567 | 1.4856 | 0.8 |
| 0.0 | 22.0 | 594 | 1.4944 | 0.8 |
| 0.0 | 23.0 | 621 | 1.5014 | 0.7778 |
| 0.0 | 24.0 | 648 | 1.5084 | 0.7778 |
| 0.0 | 25.0 | 675 | 1.5172 | 0.7778 |
| 0.0 | 26.0 | 702 | 1.5263 | 0.7778 |
| 0.0 | 27.0 | 729 | 1.5355 | 0.8 |
| 0.0 | 28.0 | 756 | 1.5446 | 0.8 |
| 0.0 | 29.0 | 783 | 1.5554 | 0.8 |
| 0.0 | 30.0 | 810 | 1.5646 | 0.8 |
| 0.0 | 31.0 | 837 | 1.5743 | 0.8 |
| 0.0 | 32.0 | 864 | 1.5840 | 0.8 |
| 0.0 | 33.0 | 891 | 1.5939 | 0.8 |
| 0.0 | 34.0 | 918 | 1.6028 | 0.8 |
| 0.0 | 35.0 | 945 | 1.6110 | 0.8 |
| 0.0 | 36.0 | 972 | 1.6204 | 0.8 |
| 0.0 | 37.0 | 999 | 1.6301 | 0.8 |
| 0.0 | 38.0 | 1026 | 1.6390 | 0.8 |
| 0.0 | 39.0 | 1053 | 1.6478 | 0.8 |
| 0.0 | 40.0 | 1080 | 1.6564 | 0.8 |
| 0.0 | 41.0 | 1107 | 1.6642 | 0.8 |
| 0.0 | 42.0 | 1134 | 1.6717 | 0.8 |
| 0.0 | 43.0 | 1161 | 1.6784 | 0.8 |
| 0.0 | 44.0 | 1188 | 1.6845 | 0.8222 |
| 0.0 | 45.0 | 1215 | 1.6898 | 0.8222 |
| 0.0 | 46.0 | 1242 | 1.6940 | 0.8222 |
| 0.0 | 47.0 | 1269 | 1.6971 | 0.8222 |
| 0.0 | 48.0 | 1296 | 1.6983 | 0.8222 |
| 0.0 | 49.0 | 1323 | 1.6984 | 0.8222 |
| 0.0 | 50.0 | 1350 | 1.6984 | 0.8222 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0