feat: add gRPC serving for @bentoml.service()

[skundu42]

What does this PR address?

bentoml serve already type-forks: legacy bentoml.legacy.Service vs @bentoml.service(). bentoml serve-grpc always called serve_grpc_production, which rejects new-style services (Service type doesn't support gRPC serving).

This adds a parallel gRPC stack for @bentoml.service(), mirroring the HTTP new-SDK path:

• Type-fork in bentoml serve-grpc and bentoml.serve(..., server_type="grpc")
• v1 unary Call + ServiceMetadata for new-style APIs
• Codec for str (text), pydantic/dict/list (JSON), np.ndarray (ndarray), and files
• Streaming, batch, and task APIs return UNIMPLEMENTED with a clear message
• serialized_bytes / pickle payloads are rejected
• Legacy 1.1 services still use serve_grpc_production

Fixes #5607
Related #4634

Before submitting:

• Does the Pull Request follow Conventional Commits specification naming?
• Does the code follow BentoML's code style, pre-commit run -a script has passed?
• Did you read through contribution guidelines and follow development guidelines?
• Did your changes require updates to the documentation? Have you updated those accordingly?
• Did you write tests to cover your changes?

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 66293f6c43

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".