Support for guided decoding for vllm backend

[Inkorak]

Is your feature request related to a problem? Please describe.
vLLM supports structured output capability with guided decoding, but the vLLM backend does not have this capability.

Describe the solution you'd like
Add Guided Decoding for vllm backend, so we can using it throught generate endpoint.

Describe alternatives you've considered
We could write our own backend modification, but it would be nice if it worked out of the box.

[tanmayv25]

Working link for guided decoding: https://docs.vllm.ai/en/latest/features/structured_outputs.html
Need to investigate what needs to be updated in vllm backend to support the feature.