Contribute vLLM wheel build or dockerfile to upstream?

[mgoin]

Hello and thank you for your great image!

I am a committer on the vLLM project and wanted to ask if you would be interested in contributing an official gh200 dockerfile to vLLM. The current dockerfile instructions/integration (https://docs.vllm.ai/en/latest/deployment/docker.html#building-for-arm64-aarch64) are minimal at best, so it would be great to have improvements.