Suggestion Description
Like nixl (https://github.com/vllm-project/vllm/blob/b569620f720d0c07bcb90bcc29dacaf0ce318d85/docker/Dockerfile#L801), could we have prebuilt wheel releases for better user experience and avoid building from source on inference framework like vLLM (vllm-project/vllm#38371)
This is ideal for reusing Mori across multiple repos as well.
Operating System
No response
GPU
No response
ROCm Component
No response
Suggestion Description
Like nixl (https://github.com/vllm-project/vllm/blob/b569620f720d0c07bcb90bcc29dacaf0ce318d85/docker/Dockerfile#L801), could we have prebuilt wheel releases for better user experience and avoid building from source on inference framework like vLLM (vllm-project/vllm#38371)
This is ideal for reusing Mori across multiple repos as well.
Operating System
No response
GPU
No response
ROCm Component
No response