Add ray-serve-vllm deployment sample

[nkwangleiGIT]

1. Add script to run LLM serving with Ray Serve
2. Add Dockerfile and RayServie yaml to prepare and deploy inference service using CR