# Load model directly
from transformers import AutoImageProcessor, AutoModelForImageClassification
processor = AutoImageProcessor.from_pretrained("mansee/swin-tiny-patch4-window7-224-img_orientation")
model = AutoModelForImageClassification.from_pretrained("mansee/swin-tiny-patch4-window7-224-img_orientation", device_map="auto")Quick Links
swin-tiny-patch4-window7-224-img_orientation
This model is a fine-tuned version of mansee/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.1069
- Accuracy: 0.9645
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: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.5605 | 1.0 | 506 | 0.3984 | 0.8341 |
| 0.3828 | 2.0 | 1013 | 0.1944 | 0.9271 |
| 0.3092 | 3.0 | 1519 | 0.1862 | 0.9339 |
| 0.3234 | 4.0 | 2026 | 0.1415 | 0.9510 |
| 0.2471 | 5.0 | 2532 | 0.1355 | 0.9517 |
| 0.251 | 6.0 | 3039 | 0.1170 | 0.9606 |
| 0.2276 | 7.0 | 3545 | 0.1136 | 0.9627 |
| 0.2182 | 8.0 | 4052 | 0.1121 | 0.9628 |
| 0.1386 | 9.0 | 4558 | 0.1116 | 0.9632 |
| 0.1466 | 9.99 | 5060 | 0.1069 | 0.9645 |
Framework versions
- Transformers 4.33.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3
- Downloads last month
- 370
Space using mansee/swin-tiny-patch4-window7-224-img_orientation 1
Evaluation results
- Accuracy on imagefolderself-reported0.964
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-classification", model="mansee/swin-tiny-patch4-window7-224-img_orientation") pipe("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/hub/parrots.png")