swinv2-tiny-patch4-window8-256-Kontur-competition-1.3K
This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.0190
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: 5e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 0.84 | 4 | 0.6834 |
| No log | 1.89 | 9 | 0.4899 |
| 0.6773 | 2.95 | 14 | 0.2752 |
| 0.6773 | 4.0 | 19 | 0.1311 |
| 0.3528 | 4.84 | 23 | 0.0812 |
| 0.3528 | 5.89 | 28 | 0.0477 |
| 0.1595 | 6.95 | 33 | 0.0289 |
| 0.1595 | 8.0 | 38 | 0.0229 |
| 0.1054 | 8.84 | 42 | 0.0213 |
| 0.1054 | 9.89 | 47 | 0.0281 |
| 0.0822 | 10.95 | 52 | 0.0190 |
| 0.0822 | 12.0 | 57 | 0.0240 |
| 0.076 | 12.63 | 60 | 0.0237 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2