Add `huggingface` model deployer support

[strickvl]

ZenML recently added support for the use of Hugging Face as a model deployer. This issue would bring support for that into mlstacks through a stack component. The component would probably be quite simple, as all that needs to happen is to define the interfaces and the output as part of the stack.yaml file. This would allow the deployer to be seamlessly imported into core ZenML and used.

The update would include:

• the core implementation
• documentation updates for all cloud providers
• any testing needed (probably minimal for this one)

[sanskriti2005]

Hi, I’m interested in working on this issue. I have experience with Python and some AI concepts, but I’m new to ZenML and Hugging Face. I’m eager to learn and would appreciate any initial guidance or resources you can provide. Could I be assigned to this issue? Thanks!

[strickvl]

This probably needs less ML / AI and more just Python + a willingness to see how mlstacks works. I'd suggest you get to know the codebase and play around with it (try the k3d provider as it's free). Happy to assign it to you and please let me know when you have any questions.

[sanskriti2005]

Alrighty, thank you!

[sanskriti2005]

@strickvl I encountered an issue while trying to install mlstacks. It seems there's a conflict with the PyYAML dependency version. mlstacks insists on installing PyYAML 6.0 version.

I installed the dependencies externally and successfully installed the latest PyYAML 6.0.1 version as well as other dependencies but when I tried to install mlstacks I ran into an error as mlstacks insists on downloading PyYAML 6.0 version.

[parthmshah1302]

I would like to work on this if it is available @strickvl

[strickvl]

@parthmshah1302 this is actually being worked on by @sanskriti2005. There is an open PR already for this request...