Skip to content

Fix memory planning error #16165

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
DrJessop wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Fix memory planning error #16165

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: 9 additions & 1 deletion backends/cadence/aot/replace_ops.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -370,15 +370,23 @@ def maybe_remove_or_replace(self, node: torch.fx.Node) -> bool:

graph = node.graph

fit_bias = is_node_with_op(bias, "get_attr") or beta == 1.0
fit_mat2 = is_node_with_op(mat2, "get_attr")

# Handle transpose: if mat2 is a transpose op, extract the original tensor
transposed_mat2 = False
if (
mat2.op == "call_function"
not fit_mat2
and mat2.op == "call_function"
and mat2.target == exir_ops.edge.aten.transpose_copy.int
):
# mat2 is already transposed, so we use the input to the transpose
mat2 = cast(torch.fx.Node, mat2.args[0])
transposed_mat2 = True
fit_mat2 = is_node_with_op(mat2, "get_attr") or alpha == 1.0

if not (fit_bias and fit_mat2):
return False

# Multiply bias by beta if needed
if beta != 1.0:
Expand Down
Loading