Add stable-diffusion model wrapper

[cmgzn]

---

name: Pull Request
about: Create a pull request

Description

add stable-diffusion model wrapper
Integrate https://github.com/AUTOMATIC1111/stable-diffusion-webui

Checklist

Please check the following items before code is ready to be reviewed.

• Code has passed all tests
• Docstrings have been added/updated in Google Style
• Documentation has been updated
• Code is ready for review

[rayrayraykk]

Please refer to the inline comments. Thank you!
Additionally, is it possible to launch the SD API server using a single, (interactive) bash script?

[rayrayraykk]

Please change name to StableDiffusionImageSynthesisWrapper

[cmgzn]

Please change name to StableDiffusionImageSynthesisWrapper

Since the SD API also provides an img2img interface, I think that both txt2img and img2img represent image synthesis, I'm concerned that naming the wrapper StableDiffusionImageSynthesisWrapper to represent text-to-image functionality might lead to confusion in the future if we decide to add an image-to-image wrapper.

[rayrayraykk]

@DavdGao Please hv a look into this, thx.

[rayrayraykk]

And please pass the UT, thx!

[cmgzn]

I've made some updates to the PR:

1. Replaced the custom request with the third-party package webuiapi. Let me know if this is suitable for the project or if you prefer a different approach.
2. I've wrapped the SD model as a callable agent service (similar to 'dashscope_services'), as I realized that directly conversing with the text2img model was not a good idea. But the dialogue agent for direct conversation with the SD model is still available and can be modified based on your feedback.

[rayrayraykk]

Could you please further provide a minimal hardware requirement for running this demo on README? And the other part looks good to me. @DavdGao Please have a look at this PR.