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 11, 03:45 UTC
[Model] Fix bagel failed to run (#31132)
Signed-off-by: wangli <wangli858794774@gmail.com>
Li Wang committed
Dec 22, 2025 at 18:15 UTC
256a33ecb4923a4bfa6e1a9adbc69f5255a5aa6d
1 file changed
+1
-1
vllm/model_executor/models/bagel.py
+1
-1
@@ -487,7 +487,7 @@ class BagelForConditionalGeneration(
487
# Split by image
488
return tuple(vision_embeds)
489
490
- def get_multimodal_embeddings(self, **kwargs: object) -> MultiModalEmbeddings:
490
+ def embed_multimodal(self, **kwargs: object) -> MultiModalEmbeddings:
491
"""Get multimodal embeddings from input."""
492
image_input = self._parse_and_validate_image_input(**kwargs)
493
if image_input is None: