vllm

feat(model): Add BitsAndBytes quantization support for Qwen3-Omni-MoE

#29896

Merged