Skip to content

[Bugfix] Fix num video tokens calculation for Qwen2-VL #13148

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
Feb 12, 2025
+5 −1
Merged

[Bugfix] Fix num video tokens calculation for Qwen2-VL #13148

Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
b6ec09b
Fix num tokens calculation for Qwen2-VL
DarkLight1337 Feb 12, 2025
1bc493e
Avoid magic number
DarkLight1337 Feb 12, 2025
2d0a58c
Update comment
DarkLight1337 Feb 12, 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
4 changes: 3 additions & 1 deletion vllm/model_executor/models/qwen2_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -831,10 +831,12 @@ def get_num_video_tokens(
num_frames: int,
image_processor: Optional[Qwen2VLImageProcessor],
) -> int:
# NOTE: Frames are padded to the next even number
# https://github.com/huggingface/transformers/blob/v4.48.3/src/transformers/models/qwen2_vl/image_processing_qwen2_vl.py#L294
_, num_video_tokens = self._get_vision_info(
image_width=image_width,
image_height=image_height,
num_frames=num_frames,
num_frames=num_frames + num_frames % 2,
image_processor=image_processor,
)
return num_video_tokens
Expand Down