Skip to content

Remove unnecessary CUDA sync of qwen image and video preprocess #22792

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Aug 13, 2025
Merged

Remove unnecessary CUDA sync of qwen image and video preprocess #22792

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
ee26070
Fix CUDA sync of qwen image and video preprocess
cyyever Aug 13, 2025
24215cf
Update vllm/model_executor/models/qwen2_5_vl.py
cyyever Aug 13, 2025
19d7685
Apply sugguestion
cyyever Aug 13, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 8 additions & 4 deletions vllm/model_executor/models/qwen2_5_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -976,10 +976,12 @@ def _process_image_input(
image_embeds = self.visual(pixel_values, grid_thw=grid_thw_list)

# Split concatenated embeddings for each image item.
# Using prod on grid_thw_list instead of grid_thw.prod avoids CUDA sync
merge_size = self.visual.spatial_merge_size
sizes = grid_thw.prod(-1) // merge_size // merge_size
sizes = (torch.tensor(grid_thw_list, dtype=torch.long).prod(-1) //
(merge_size * merge_size)).tolist()

return image_embeds.split(sizes.tolist())
return image_embeds.split(sizes)

def _process_video_input(
self,
Expand All @@ -998,9 +1000,11 @@ def _process_video_input(

# Split concatenated embeddings for each video item.
merge_size = self.visual.spatial_merge_size
sizes = grid_thw.prod(-1) // merge_size // merge_size
# Using prod on grid_thw_list instead of grid_thw.prod avoids CUDA sync
sizes = (torch.tensor(grid_thw_list, dtype=torch.long).prod(-1) //
(merge_size * merge_size)).tolist()

return video_embeds.split(sizes.tolist())
return video_embeds.split(sizes)

def _parse_and_validate_multimodal_inputs(self, **kwargs: object) -> dict:
mm_input_by_modality = {}
Expand Down