OwLite is designed to support a wide spectrum of AI models, underscoring its versatility and broad applicability.
The models in our documentation serve as verified examples, showcasing 'Example code ready' and 'Recommended confirmed' status to demonstrate proven compatibility with OwLite.
While OwLite also supports compression for unlisted models, users may encounter particular challenges in these scenarios. Our Recommendation algorithm is also designed to be applicable across a diverse set of models, although there may be exceptional cases.
If you need help with unlisted models, we'd appreciate your feedback to improve OwLite!
TASK | Model name | Example code | Recommended confirmed |
---|---|---|---|
🟠 IMAGE CLASSIFICATION | ResNet18 | ✅ | ✅ |
🟠 IMAGE CLASSIFICATION | ResNet50 | ✅ | ✅ |
🟠 IMAGE CLASSIFICATION | MobileNet-V2 | ✅ | ✅ |
🟠 IMAGE CLASSIFICATION | MobileNet-V3-Large | ✅ | ✅ |
🟠 IMAGE CLASSIFICATION | EfficientNet-B0 | ✅ | ✅ |
🟠 IMAGE CLASSIFICATION | EfficientNet-V2-S | ✅ | ✅ |
🟠 IMAGE CLASSIFICATION | Swin-B | ✅ | ✅ |
🟠 IMAGE-SEGMENTATION | Deeplabv3_Resnet50 | ✅ | ✅ |
🔴 OBJECT-DETECTION | YOLOv5 | ✅ | ✅ |
🔴 OBJECT-DETECTION | YOLOv6 | ✅ | ✅ |
🔴 OBJECT-DETECTION | YOLOv8 | ✅ | ✅ |
🔴 OBJECT-DETECTION | YOLOX | ✅ | ✅ |
🔵 TEXT-CLASSIFICATION | BERT-base | ✅ | ✅ |