search

jina-ai / clip-as-service

🏄 Scalable embedding, reasoning, ranking for images and sentences with CLIP
https://clip-as-service.jina.ai
Other
12.4k stars 2.07k forks source link

Compatibility with OWL-ViT #891

Open NOT-HAL9000 opened 1 year ago

NOT-HAL9000 commented 1 year ago

Would it be possible to get this working with OWL-ViT? https://segments.ai/blog/zero-shot-object-detection-with-owl-vit

shubham-goel-zefr commented 1 year ago

You can convert the OWL-ViT weights to onnx using torch.export and then supply the path manually in the jina service (in case no one picks this up).