[Perf] VoxCPM2: Speedup by manual CUDA Graph capture for scaffold/residual forward#2803

hsliuustc0106 merged 2 commits into

vllm-project:mainvllm-project/vllm-omni:mainfrom

Sy0307:perf/voxcpm2-streaming-vaeSy0307/vllm-omni:perf/voxcpm2-streaming-vaeCopy head branch name to clipboard