Skip to content

Remove Redundant Assignment in Qwen3_VisionPatchMerger #25224

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 2 commits into from
Sep 19, 2025
Merged

Remove Redundant Assignment in Qwen3_VisionPatchMerger #25224

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
033620b
opt code
Sep 19, 2025
3339176
Merge branch 'main' into qwen3_vl_opt
ywang96 Sep 19, 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: 1 addition & 3 deletions vllm/model_executor/models/qwen3_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -223,9 +223,7 @@ def __init__(

if norm_layer is None:
norm_layer = partial(nn.LayerNorm, eps=1e-6)
self.use_postshuffle_norm = use_postshuffle_norm
self.norm = norm_layer(
self.hidden_size if use_postshuffle_norm else context_dim)
self.norm = norm_layer(context_dim)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

critical

This change introduces a bug. By removing the assignment to self.use_postshuffle_norm, the forward method will raise an AttributeError because it relies on this attribute.

Furthermore, the normalization layer self.norm is now always initialized with context_dim. However, when use_postshuffle_norm is True, it should be initialized with self.hidden_size to match the tensor shape in the forward method. The current implementation will cause a shape mismatch error.

To fix this, you should restore the assignment and the conditional logic for the normalization dimension. Here is a suggested correction that keeps the logic correct while improving readability slightly.

        self.use_postshuffle_norm = use_postshuffle_norm
        norm_dim = self.hidden_size if use_postshuffle_norm else context_dim
        self.norm = norm_layer(norm_dim)

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

self.use_postshuffle_norm is initialized in L220 and context_dim will be equal to self.hidden_size if self.use_postshuffle_norm is true in L221

self.linear_fc1 = ColumnParallelLinear(self.hidden_size,
self.hidden_size,
bias=True,
Expand Down