@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / d8218b1ee7
Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

[Bugfix] Propagate ImportError from load_audio_pyav when vllm[audio] … (#44750)

Signed-off-by: littlecircle0730 <littlecircle0730@gmail.com> Signed-off-by: Hsiao-Yuan Chen <hy.c@Hsiao-YuandeMacBook-Pro.local> Co-authored-by: Hsiao-Yuan Chen <hy.c@Hsiao-YuandeMacBook-Pro.local>

littlecircle0730 committed Jun 8, 2026 at 21:24 UTC d8218b1ee7db3e7582e0151e942d17963a1e3e13
1 file changed +1 -1
vllm/multimodal/media/audio.py
+1 -1
@@ -91,7 +91,7 @@ def load_audio_pyav(
91 chunks.append(out_frame.to_ndarray())
92 else:
93 chunks.append(frame.to_ndarray())
94 - except ValueError:
94 + except (ValueError, ImportError):
95 raise
96 except Exception as e:
97 raise ValueError(