smids_10x_deit_small_sgd_00001_fold4
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: 0.9309
- Accuracy: 0.5667
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 |
|---|---|---|---|---|
| 1.049 | 1.0 | 750 | 1.0657 | 0.4217 |
| 1.0271 | 2.0 | 1500 | 1.0619 | 0.4233 |
| 1.0309 | 3.0 | 2250 | 1.0577 | 0.4233 |
| 1.0685 | 4.0 | 3000 | 1.0531 | 0.4233 |
| 1.0213 | 5.0 | 3750 | 1.0484 | 0.425 |
| 1.0075 | 6.0 | 4500 | 1.0438 | 0.4267 |
| 1.0135 | 7.0 | 5250 | 1.0390 | 0.4283 |
| 1.0193 | 8.0 | 6000 | 1.0343 | 0.43 |
| 1.0172 | 9.0 | 6750 | 1.0296 | 0.4383 |
| 0.995 | 10.0 | 7500 | 1.0249 | 0.4417 |
| 0.9861 | 11.0 | 8250 | 1.0204 | 0.4467 |
| 0.9925 | 12.0 | 9000 | 1.0158 | 0.4533 |
| 0.9841 | 13.0 | 9750 | 1.0115 | 0.465 |
| 0.9738 | 14.0 | 10500 | 1.0072 | 0.4733 |
| 0.9779 | 15.0 | 11250 | 1.0030 | 0.4783 |
| 0.9393 | 16.0 | 12000 | 0.9988 | 0.485 |
| 0.968 | 17.0 | 12750 | 0.9949 | 0.485 |
| 0.9542 | 18.0 | 13500 | 0.9909 | 0.4883 |
| 0.9456 | 19.0 | 14250 | 0.9871 | 0.4917 |
| 0.9805 | 20.0 | 15000 | 0.9834 | 0.4967 |
| 0.9272 | 21.0 | 15750 | 0.9798 | 0.5 |
| 0.9402 | 22.0 | 16500 | 0.9763 | 0.5083 |
| 0.9463 | 23.0 | 17250 | 0.9729 | 0.5133 |
| 0.9349 | 24.0 | 18000 | 0.9697 | 0.515 |
| 0.9212 | 25.0 | 18750 | 0.9666 | 0.5167 |
| 0.9115 | 26.0 | 19500 | 0.9636 | 0.5183 |
| 0.9201 | 27.0 | 20250 | 0.9607 | 0.5217 |
| 0.9475 | 28.0 | 21000 | 0.9580 | 0.525 |
| 0.9135 | 29.0 | 21750 | 0.9554 | 0.5267 |
| 0.9341 | 30.0 | 22500 | 0.9529 | 0.53 |
| 0.9173 | 31.0 | 23250 | 0.9505 | 0.5317 |
| 0.9276 | 32.0 | 24000 | 0.9483 | 0.535 |
| 0.9211 | 33.0 | 24750 | 0.9462 | 0.5417 |
| 0.9232 | 34.0 | 25500 | 0.9443 | 0.5467 |
| 0.9171 | 35.0 | 26250 | 0.9425 | 0.5483 |
| 0.9007 | 36.0 | 27000 | 0.9408 | 0.5483 |
| 0.9143 | 37.0 | 27750 | 0.9393 | 0.555 |
| 0.8916 | 38.0 | 28500 | 0.9379 | 0.5567 |
| 0.8951 | 39.0 | 29250 | 0.9366 | 0.5567 |
| 0.9014 | 40.0 | 30000 | 0.9355 | 0.5567 |
| 0.8889 | 41.0 | 30750 | 0.9345 | 0.5583 |
| 0.8953 | 42.0 | 31500 | 0.9336 | 0.5583 |
| 0.9154 | 43.0 | 32250 | 0.9329 | 0.5583 |
| 0.8836 | 44.0 | 33000 | 0.9323 | 0.5633 |
| 0.8961 | 45.0 | 33750 | 0.9318 | 0.5667 |
| 0.8837 | 46.0 | 34500 | 0.9314 | 0.5667 |
| 0.8621 | 47.0 | 35250 | 0.9311 | 0.5667 |
| 0.8982 | 48.0 | 36000 | 0.9310 | 0.5667 |
| 0.8793 | 49.0 | 36750 | 0.9309 | 0.5667 |
| 0.8813 | 50.0 | 37500 | 0.9309 | 0.5667 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2