Skip to content

Forward fix Fix batch norm partitioning with Conv3d (#13696) #14170

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 1 commit into from
Sep 10, 2025
Merged

Forward fix Fix batch norm partitioning with Conv3d (#13696) #14170

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
10 changes: 8 additions & 2 deletions backends/xnnpack/_passes/fuse_batch_norm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -115,8 +115,14 @@ def can_fuse(
return False

# Check the rank of the convolutution input - only Conv1d and 2d are supported.
if is_conv and len(input_node.args[0].meta["val"].shape) not in (3, 4):
return False
if is_conv:
conv_input = input_node.args[0]
if (
not isinstance(conv_input, torch.fx.Node)
or "val" not in conv_input.meta
or len(conv_input.meta["val"].shape) not in (3, 4)
):
return False

return True

Expand Down
2 changes: 1 addition & 1 deletion backends/xnnpack/test/passes/test_batch_norm_fusion.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -60,7 +60,7 @@ def __init__(
self,
in_features: int,
out_features: int,
kernel_size: Tuple[int, int],
kernel_size: Tuple[int, int, int],
):
super().__init__()
op = torch.nn.Conv3d
Expand Down
Loading