feat: split core and extra images with additional backends

[9876691]

The image is so large that I can't run a Github action as it runs out of space.

quay.io/go-skynet/local-ai v1.30.0  ba9a2b1bff11   2 weeks ago   14GB

Using dive I can see the image has Cuda installed and this takes up a lot of space.

Possible solution

Have a CPU only image with no Cuda?

[mudler]

fair point, the images got larger as backends were added, and I guess it will just get fatter due to #1144 .

We could - at least - split the images to a core and an extra version, which comes with all the python backends as well.

[Aisuko]

Can we split the backend by type? like C++, Rust and Python. These backends can be plugins

[localai-bot]

⚠️⚠️⚠️⚠️⚠️

Hi! I'm a bot running with LocalAI ( a crazy experiment of @mudler ) - please beware that I might hallucinate sometimes!

but.... I can also be funny or helpful 😸 and I can provide generally speaking good tips or places where to look after in the documentation or in the code based on what you wrote in the issue.

Don't engage in conversation with me,
I don't support (yet) replying!

⚠️⚠️⚠️⚠️⚠️

Yes, that is a possible solution. You could create a new Docker image without CUDA and upload it to quay.io. Then you can use that image in your docker-compose file instead of the current one. This will reduce the size of the image and allow you to run Github actions without running out of space.

Sources:

• https://github.com/go-skynet/LocalAI/tree/master/docker-compose.yaml
• https://localai.io/basics/getting_started/index.html
• https://localai.io/howtos/easy-setup-docker-gpu/index.html

[mudler]

this is still much wanted as today - we have now a flag to build smaller images (IMAGE_TYPE=core) but needs to be added to the pipelines

[mudler]

#1309 introduces -core images, let's check what builds out of master and see if we can improve the size of those