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 13, 08:15 UTC
[Bugfix] Fix FPS value type for Qwen2.5-Omni video processing (#28630)
Signed-off-by: Lin, Fanli <fanli.lin@intel.com>
Fanli Lin committed
Nov 13, 2025 at 21:06 UTC
c9fe6abe7c0b03d552420edd63c6c678ed683dea
1 file changed
+1
-1
examples/offline_inference/vision_language.py
+1
-1
@@ -1536,7 +1536,7 @@ def run_qwen2_5_omni(questions: list[str], modality: str):
1536
mm_processor_kwargs={
1537
"min_pixels": 28 * 28,
1538
"max_pixels": 1280 * 28 * 28,
1539
- "fps": [1],
1539
+ "fps": 1,
1540
},
1541
limit_mm_per_prompt={modality: 1},
1542
)