A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 2, 22:00 UTC
[Bugfix][Frontend] Fix input_audio parsing when uuid is present (#43414)
Signed-off-by: ffggs <314137448@qq.com> Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>
ffggs committed
May 24, 2026 at 00:03 UTC
d8b385b7eac4724425f75bfe7a7ae3f477dd8d5d
1 file changed
+1
-1
vllm/entrypoints/chat_utils.py
+1
-1
@@ -1517,7 +1517,7 @@ def _parse_chat_message_content_mm_part(
1517
audio_url = audio_url.get("url", None)
1518
return "audio_url", audio_url
1519
if part.get("input_audio") is not None:
1520
- input_audio_params = cast(dict[str, str], part)
1520
+ input_audio_params = _InputAudioParser(part).get("input_audio", None)
1521
return "input_audio", input_audio_params
1522
if "video_url" in part:
1523
video_params = cast(CustomChatCompletionContentSimpleVideoParam, part)