swin-tiny-patch4-window7-224-finetuned-azure-poc-img-classification
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: 0.3382
- Accuracy: 0.8562
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: 128
- eval_batch_size: 128
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 512
- optimizer: Use 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: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 5.3507 | 1.0 | 14 | 0.8673 | 0.6557 |
| 3.1899 | 2.0 | 28 | 0.4795 | 0.8184 |
| 1.5944 | 3.0 | 42 | 0.3992 | 0.8436 |
| 1.3952 | 4.0 | 56 | 0.3741 | 0.8499 |
| 1.1452 | 5.0 | 70 | 0.3601 | 0.8575 |
| 1.0478 | 6.0 | 84 | 0.3470 | 0.8562 |
| 1.0049 | 7.0 | 98 | 0.3360 | 0.8562 |
| 0.9361 | 8.0 | 112 | 0.3318 | 0.8525 |
| 0.9477 | 9.0 | 126 | 0.3297 | 0.8512 |
| 0.8568 | 10.0 | 140 | 0.3382 | 0.8562 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.0