swin-tiny-patch4-window7-224-finetuned-eurosat
This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.4495
- Accuracy: 0.858
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: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 30
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.0489 | 1.0 | 211 | 1.0636 | 0.4417 |
| 0.8338 | 2.0 | 422 | 0.7686 | 0.6493 |
| 0.6886 | 3.0 | 633 | 0.6770 | 0.7043 |
| 0.6053 | 4.0 | 844 | 0.5288 | 0.7857 |
| 0.5038 | 5.0 | 1055 | 0.5047 | 0.7933 |
| 0.4413 | 6.0 | 1266 | 0.4256 | 0.823 |
| 0.3967 | 7.0 | 1477 | 0.4226 | 0.8363 |
| 0.3301 | 8.0 | 1688 | 0.4969 | 0.81 |
| 0.3331 | 9.0 | 1899 | 0.3893 | 0.8537 |
| 0.3213 | 10.0 | 2110 | 0.4314 | 0.8523 |
| 0.2594 | 11.0 | 2321 | 0.4389 | 0.8337 |
| 0.2161 | 12.0 | 2532 | 0.3911 | 0.8637 |
| 0.244 | 13.0 | 2743 | 0.3854 | 0.8533 |
| 0.1849 | 14.0 | 2954 | 0.4328 | 0.8633 |
| 0.1787 | 15.0 | 3165 | 0.4495 | 0.858 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2