feat: unify and propagate CMAKE_ARGS to GGML-based backends

[mudler]

Description

This pull request centralizes CMAKE_ARGS composition as such is shared between backends that are ggml-based. llama.cpp cmake args can be used for instance with bark.cpp and stable-diffusion.cpp (ggml based). This aims to enable cuda and hipblas support on bark.cpp and stablediffusion.cpp (ggml variant).

For now this doesn't aim to be smart and share this in a common way (maybe using cmake, or a makefile that is called by both to generate the cmake args). The attempt of this pr is to understand any changes that might be required by enabling the flags for the respective backends. I'm not sure that bark and stablediffusion currently linking processes are correct in term of GPU support.

Notes for Reviewers

Signed commits

• [ ] Yes, I signed my commits.

[netlify[bot]]

✅ Deploy Preview for localai ready!

Name	Link
🔨 Latest commit	894a30296a14c1c14549db05a0177f0b2a448c65
🔍 Latest deploy log	https://app.netlify.com/sites/localai/deploys/6759fe760a6baa000818c441
😎 Deploy Preview	https://deploy-preview-4367--localai.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[github-actions[bot]]

This PR is stale because it has been open 90 days with no activity. Remove stale label or comment or this will be closed in 10 days.

[github-actions[bot]]

This PR was closed because it has been stalled for 10 days with no activity.