swin-small-patch4-window7-224-finetuned-eurosat
This model is a fine-tuned version of microsoft/swin-small-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.9061
- Accuracy: 0.7232
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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 12
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.1265 | 1.0 | 77 | 1.1283 | 0.6531 |
| 1.0442 | 2.0 | 154 | 1.0860 | 0.6568 |
| 1.0043 | 3.0 | 231 | 1.0477 | 0.6605 |
| 0.9986 | 4.0 | 308 | 1.0174 | 0.6900 |
| 1.0202 | 5.0 | 385 | 0.9889 | 0.6863 |
| 0.9499 | 6.0 | 462 | 0.9659 | 0.7011 |
| 0.9138 | 7.0 | 539 | 0.9453 | 0.7122 |
| 0.9211 | 8.0 | 616 | 0.9305 | 0.7122 |
| 0.9263 | 9.0 | 693 | 0.9194 | 0.7196 |
| 0.8678 | 10.0 | 770 | 0.9114 | 0.7232 |
| 0.9109 | 11.0 | 847 | 0.9071 | 0.7196 |
| 0.9316 | 12.0 | 924 | 0.9061 | 0.7232 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.6.0
- Tokenizers 0.21.0