hushem_1x_deit_tiny_sgd_0001_fold5
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: 1.5523
- Accuracy: 0.2439
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 |
|---|---|---|---|---|
| No log | 1.0 | 6 | 1.7547 | 0.2439 |
| 1.7078 | 2.0 | 12 | 1.7422 | 0.2439 |
| 1.7078 | 3.0 | 18 | 1.7303 | 0.2439 |
| 1.6827 | 4.0 | 24 | 1.7187 | 0.2439 |
| 1.6676 | 5.0 | 30 | 1.7076 | 0.2439 |
| 1.6676 | 6.0 | 36 | 1.6970 | 0.2439 |
| 1.6669 | 7.0 | 42 | 1.6882 | 0.2439 |
| 1.6669 | 8.0 | 48 | 1.6793 | 0.2439 |
| 1.5935 | 9.0 | 54 | 1.6701 | 0.2439 |
| 1.6316 | 10.0 | 60 | 1.6617 | 0.2439 |
| 1.6316 | 11.0 | 66 | 1.6538 | 0.2439 |
| 1.6324 | 12.0 | 72 | 1.6460 | 0.2439 |
| 1.6324 | 13.0 | 78 | 1.6387 | 0.2439 |
| 1.5842 | 14.0 | 84 | 1.6318 | 0.2439 |
| 1.5897 | 15.0 | 90 | 1.6256 | 0.2439 |
| 1.5897 | 16.0 | 96 | 1.6199 | 0.2439 |
| 1.5943 | 17.0 | 102 | 1.6144 | 0.2439 |
| 1.5943 | 18.0 | 108 | 1.6092 | 0.2195 |
| 1.5586 | 19.0 | 114 | 1.6040 | 0.2195 |
| 1.5924 | 20.0 | 120 | 1.5990 | 0.2195 |
| 1.5924 | 21.0 | 126 | 1.5945 | 0.2195 |
| 1.5676 | 22.0 | 132 | 1.5902 | 0.2195 |
| 1.5676 | 23.0 | 138 | 1.5862 | 0.2195 |
| 1.5352 | 24.0 | 144 | 1.5823 | 0.2195 |
| 1.5842 | 25.0 | 150 | 1.5786 | 0.2195 |
| 1.5842 | 26.0 | 156 | 1.5752 | 0.2195 |
| 1.5461 | 27.0 | 162 | 1.5723 | 0.2195 |
| 1.5461 | 28.0 | 168 | 1.5695 | 0.2195 |
| 1.551 | 29.0 | 174 | 1.5671 | 0.2439 |
| 1.5549 | 30.0 | 180 | 1.5649 | 0.2439 |
| 1.5549 | 31.0 | 186 | 1.5628 | 0.2439 |
| 1.5532 | 32.0 | 192 | 1.5610 | 0.2439 |
| 1.5532 | 33.0 | 198 | 1.5594 | 0.2439 |
| 1.5006 | 34.0 | 204 | 1.5578 | 0.2439 |
| 1.5134 | 35.0 | 210 | 1.5565 | 0.2439 |
| 1.5134 | 36.0 | 216 | 1.5553 | 0.2439 |
| 1.5386 | 37.0 | 222 | 1.5543 | 0.2439 |
| 1.5386 | 38.0 | 228 | 1.5536 | 0.2439 |
| 1.5372 | 39.0 | 234 | 1.5530 | 0.2439 |
| 1.528 | 40.0 | 240 | 1.5526 | 0.2439 |
| 1.528 | 41.0 | 246 | 1.5524 | 0.2439 |
| 1.5555 | 42.0 | 252 | 1.5523 | 0.2439 |
| 1.5555 | 43.0 | 258 | 1.5523 | 0.2439 |
| 1.509 | 44.0 | 264 | 1.5523 | 0.2439 |
| 1.5379 | 45.0 | 270 | 1.5523 | 0.2439 |
| 1.5379 | 46.0 | 276 | 1.5523 | 0.2439 |
| 1.5588 | 47.0 | 282 | 1.5523 | 0.2439 |
| 1.5588 | 48.0 | 288 | 1.5523 | 0.2439 |
| 1.509 | 49.0 | 294 | 1.5523 | 0.2439 |
| 1.5414 | 50.0 | 300 | 1.5523 | 0.2439 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1