### What this PR does / why we need it? Add Qwen3Next support in main ### Does this PR introduce _any_ user-facing change? ### How was this patch tested? - vLLM version: v0.11.2 - vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.2 --------- Signed-off-by: SunnyLee219 <3294305115@qq.com>