fix: use ref_audio path for uploaded custom voices in voxtral TTS inference (#2479)#2547

passionworkeer wants to merge 1 commit into

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

passionworkeer:fix-uploaded-custom-voice-ref-audiopassionworkeer/vllm-omni:fix-uploaded-custom-voice-ref-audioCopy head branch name to clipboard