Add llama-cpp-python

[acon96]

Adds wheels for llama-cpp-python to allow running Llama AI models as part of custom integrations.

[home-assistant]

Hi @acon96

It seems you haven't yet signed a CLA. Please do so here.

Once you do that we will be able to review and accept this pull request.

Thanks!

[home-assistant]

Please take a look at the requested changes, and use the Ready for review button when you are done, thanks 👍

Learn more about our pull request process.

Stale

[frenck]

It seems that upstream already publishes musl linux wheels, why does it need to be added here as well?

[acon96]

Can you provide the link to the upstream musl wheels? PyPi still only provides a source tarball and no binaries at all.

[frenck]

Hmm it seems that I looked wrongly. They are indeed not on Pypi. Maybe I've looked at the wrong package.

Have you contacted the author to ask if they are willing to publish Wheels?

[acon96]

Looks like they're still working on it and don't plan to publish them to PyPi. Not sure if that is compatible with how Home Assistant installs packages so I'll figure something else out. Thanks

[acon96]

hmm. that definitely didn't exist a month ago. thanks for the heads up

[abetlen]

Hey @acon96 @frenck with abetlen/llama-cpp-python#1247 now closed it's possible to install pre-built binary llama-cpp-python wheels using --extra-index-url similar to say a PyTorch installation. This method is compatible with requirements.txt files ie:

# requirements.txt
--extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu
llama-cpp-python

This specifies the basic cpu wheels for llama-cpp-python and allows you to install it without a compiler toolchain.