feat(helm/raycluster): pass through annotations and podAnnotations (#980)

The `annotations` and `podAnnotations` values already exist on
`modelSpec` and are wired into the standard (non-Ray) vLLM Deployment
via `deployment-vllm-multi.yaml`, but the RayCluster template never
consumed them. As a result, users running engines through KubeRay had
no way to attach pod-level annotations (e.g. Multus
`k8s.v1.cni.cncf.io/networks`) or object-level annotations to their
head/worker pods.

Mirror the Deployment pattern in `ray-cluster.yaml`:
- `modelSpec.annotations` -> RayCluster object metadata
- `modelSpec.podAnnotations` -> head and worker pod template metadata

No new values are introduced; this only completes the existing
passthrough for the KubeRay path. Adds a helm-unittest case asserting
both annotation kinds render on the RayCluster.

Closes #962

Signed-off-by: Tai An <antai12232931@outlook.com>
Co-authored-by: Rui Zhang <51696593+ruizhang0101@users.noreply.github.com>