[Test][Mamba] add e2e tests for Mamba/hybrid models (#14981)

### What this PR does / why we need it?

This PR adds end-to-end tests for Mamba and hybrid-architecture models.
This test case runs inference for Qwen3.5‑27B under both `enforce_eager`
and CUDAGraph modes.

### How was this patch tested?

The newly added end-to-end tests passed locally.

- vLLM version: v0.27.1
- vLLM main:
https://github.com/vllm-project/vllm/commit/ba07e4a48fc951300d97eb506217dd530583dea3

---------

Signed-off-by: Lethobenthos20 <3205604419@qq.com>