"ViT-L-14@336px" model name doesn't work with CLIPTorchEncoder.

[delgermurun]

ViT-L-14@336px model name mentioned in the docs, but it doesn't work anymore.

    uses: jinahub+docker://CLIPTorchEncoder/latest-gpu
    uses_with:
      name: ViT-L-14@336px

---

But ViT-L-14-336::openai works.

    uses: jinahub+docker://CLIPTorchEncoder/latest-gpu
    uses_with:
      name: ViT-L-14-336::openai

[ZiniuYu]

Hi @delgermurun , it is a naming convention. We used to only support OpenAI models and they called it ViT-L/14@336px (left slash). After we switch to OpenCLIP models, we changed to ViT-L-14-336::openai. We kept the old names for backward compatibility. But I do think we need to clarify this or have higher fault tolerance on those names. Thanks for pointing this out!

[delgermurun]

below image is from https://hub.jina.ai/executor/gzpbl8jh

here even old name is wrong.