Feature Request: Add a docker target for the llama app

[kannon92]

Prerequisites

• I am running the latest code. Mention the version if possible as well.
• I carefully followed the README.md.
• I searched using keywords relevant to my issue to make sure that I am creating a new issue that is not already open (or closed).
• I reviewed the Discussions, and have a new and useful enhancement to share.

Feature Description

The llama app command is really exciting and its neat to have the ability to run with different models.

It would be great to publish Dockerfiles for this as you all do for server, cli, full, etc.

Motivation

Dockerfiles are a great package tool for Kubernetes and provide reproducible environments.

docker run -p 8080:8080 -v ~/.cache/huggingface/hub:/root/.cache/huggingface/hub llama-app serve --host 0.0.0.0

Possible Implementation

I tested this out with .devops/cpu.Dockerfile.

### App, unified binary with all subcommands
FROM base AS app

COPY --from=build /app/full/llama /app

WORKDIR /app

ENTRYPOINT [ "/app/llama" ]

I then was able to run docker run -p 8080:8080 -v ~/.cache/huggingface/hub:/root/.cache/huggingface/hub llama-app serve --host 0.0.0.0 and have this running in a container while I switch to different models.

If this interests the community, I'm happy to add support for this to all our Dockerfiles and update the docker docs to mention this approach.