Add CLIP resources

[eenzeenee]

What does this PR do?

Adds resources of CLIP according to this issue

Part of #20055

Before submitting

• [x ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline,
  Pull Request section?
• Was this discussed/approved via a Github issue or the forum? Please add a link
  to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

@stevhliu, @jungnerd, @wonhyeongseo may you please review this PR?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

[jungnerd]

LGTM! Thanks for adding resources for CLIP.
By the way, please fix Part of #20555 to Part of #20055 since 20055 is the issue of Model resources contribution.

[eenzeenee]

Thank you for reviewing!! I fixed it!

[stevhliu]

Thanks for your contribution! 🤗

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint.

[NielsRogge]

Thanks for contributing resources! It's very helpful, however, a lot of resources here are a bit random.

• assigning the "text-to-image" pipeline tag to CLIP is a bit weird since CLIP is only a text encoder, vision encoder model. It can't be used for generating images using text. It is however used as a building block in Stable Diffusion for instance to condition the model on text prompts.
• the "deploy" section includes several resources which are not about deployment. Deployment is about optimizing the model for inference, using tools like ONNX, 🤗 Optimum, quantization, etc.
• the "inference" section includes a notebook about... explainability. Then it makes more sense to include this in a separate "explainability" section. Inference is about showcasing predictions with the model

Some relevant resources for CLIP include:

• fine-tuning CLIP => HF provides a training script for this: https://github.com/huggingface/transformers/tree/main/examples/pytorch/contrastive-image-text.
• fine-tuning CLIP on additional (image, text) pairs: https://huggingface.co/blog/fine-tune-clip-rsicd.

[stevhliu]

Thanks for fixing, can you also include the training script and blog post @NielsRogge linked to?

[eenzeenee]

Thanks for fixing, can you also include the training script and blog post @NielsRogge linked to?

Sorry for the late reply. It seems to appear in 86 and 87 lines. Do you want me to change the description to something else?

[stevhliu]

Do you want me to change the description to something else?

Should be good then! 👍

[stevhliu]

A few more minor nits sorry! 😅