feat: add a simple way to convert Hugging Face model to GGUF

[Helloyunho]

Note: Join Jan's Discord server and go to roadmap -> Jan can convert Hugging Face models to GGUF for more frequent updates.

Describe Your Changes

• Add a way to convert Hugging Face models to GGUF in Jan
• (Hopefully) make decent metadata from Hugging Face's model card

Fixes Issues

• None

Self Checklist

• [ ] Added relevant comments, esp in complex areas
  • TODO
• [ ] Updated docs (for bug fixes / features)
  • TODO
• [ ] Created issues for follow-up changes or refactoring needed

TODO (yeah I just made this up)

• All of the above
• Have a better UI (we really need it, the current UI that I made is um, filthy)
• Alert when the user doesn't have Python (check below)
• Build and bundle llama.cpp's quantize binary(like we do for nitro)

Things to know (again I added this)

• User should have Python installed locally
  • Long explanation: This change uses llama.cpp's convert script to convert models, and since the script is written in Python, we need it to run the script.
• For easy debugging, it doesn't remove the model directory when the conversion has failed. Later we should uncomment it before publishing.