Skip to content

Remove over-strict softmax mask divisibility assert #19903

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

Open
ethansfng wants to merge 1 commit into
base: main
Choose a base branch
from
+0 −3
Open

Remove over-strict softmax mask divisibility assert #19903

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
3 changes: 0 additions & 3 deletions backends/cadence/aot/quantizer/patterns.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1092,9 +1092,6 @@ def fuse(self, gm: fx.GraphModule, anchor_node: fx.Node) -> fx.Node | None:
return None
mask_shape = list(mask_shape)
# Softmax mask is packed 16 elements per int32 word.
assert (
mask_shape[-1] % 16 == 0
), f"Softmax mask dimension must be divisible by 16, got {mask_shape[-1]}"
mask_shape[-1] = mask_shape[-1] // 16
mask_tensor = insert_node_with_meta(
gm,
Expand Down
Loading