hushem_1x_deit_tiny_rms_0001_fold4
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: 2.3903
- Accuracy: 0.5714
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 | 1.5699 | 0.2619 |
| 2.0801 | 2.0 | 12 | 1.5693 | 0.2381 |
| 2.0801 | 3.0 | 18 | 1.6087 | 0.2619 |
| 1.5352 | 4.0 | 24 | 1.4372 | 0.2619 |
| 1.4323 | 5.0 | 30 | 1.3212 | 0.3095 |
| 1.4323 | 6.0 | 36 | 1.3803 | 0.2381 |
| 1.3894 | 7.0 | 42 | 1.4606 | 0.4524 |
| 1.3894 | 8.0 | 48 | 1.5543 | 0.2619 |
| 1.294 | 9.0 | 54 | 1.1365 | 0.5 |
| 1.1627 | 10.0 | 60 | 1.3219 | 0.3571 |
| 1.1627 | 11.0 | 66 | 1.0508 | 0.5714 |
| 1.0159 | 12.0 | 72 | 1.0736 | 0.5 |
| 1.0159 | 13.0 | 78 | 1.6175 | 0.3571 |
| 0.8051 | 14.0 | 84 | 1.4409 | 0.4524 |
| 0.5869 | 15.0 | 90 | 2.1188 | 0.4286 |
| 0.5869 | 16.0 | 96 | 1.8546 | 0.5476 |
| 0.3044 | 17.0 | 102 | 1.7485 | 0.5 |
| 0.3044 | 18.0 | 108 | 1.6544 | 0.5476 |
| 0.2005 | 19.0 | 114 | 1.7817 | 0.5714 |
| 0.0634 | 20.0 | 120 | 2.6836 | 0.5238 |
| 0.0634 | 21.0 | 126 | 2.3476 | 0.5714 |
| 0.0488 | 22.0 | 132 | 2.3551 | 0.5476 |
| 0.0488 | 23.0 | 138 | 2.4123 | 0.5714 |
| 0.0014 | 24.0 | 144 | 2.3855 | 0.5714 |
| 0.0006 | 25.0 | 150 | 2.3709 | 0.5714 |
| 0.0006 | 26.0 | 156 | 2.3623 | 0.5714 |
| 0.0004 | 27.0 | 162 | 2.3621 | 0.5714 |
| 0.0004 | 28.0 | 168 | 2.3646 | 0.5952 |
| 0.0003 | 29.0 | 174 | 2.3639 | 0.5952 |
| 0.0003 | 30.0 | 180 | 2.3665 | 0.5952 |
| 0.0003 | 31.0 | 186 | 2.3692 | 0.5952 |
| 0.0002 | 32.0 | 192 | 2.3723 | 0.5952 |
| 0.0002 | 33.0 | 198 | 2.3750 | 0.5952 |
| 0.0002 | 34.0 | 204 | 2.3777 | 0.5714 |
| 0.0002 | 35.0 | 210 | 2.3806 | 0.5714 |
| 0.0002 | 36.0 | 216 | 2.3834 | 0.5714 |
| 0.0002 | 37.0 | 222 | 2.3855 | 0.5714 |
| 0.0002 | 38.0 | 228 | 2.3872 | 0.5714 |
| 0.0001 | 39.0 | 234 | 2.3885 | 0.5714 |
| 0.0001 | 40.0 | 240 | 2.3895 | 0.5714 |
| 0.0001 | 41.0 | 246 | 2.3902 | 0.5714 |
| 0.0001 | 42.0 | 252 | 2.3903 | 0.5714 |
| 0.0001 | 43.0 | 258 | 2.3903 | 0.5714 |
| 0.0001 | 44.0 | 264 | 2.3903 | 0.5714 |
| 0.0001 | 45.0 | 270 | 2.3903 | 0.5714 |
| 0.0001 | 46.0 | 276 | 2.3903 | 0.5714 |
| 0.0001 | 47.0 | 282 | 2.3903 | 0.5714 |
| 0.0001 | 48.0 | 288 | 2.3903 | 0.5714 |
| 0.0001 | 49.0 | 294 | 2.3903 | 0.5714 |
| 0.0001 | 50.0 | 300 | 2.3903 | 0.5714 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1