Skip to content

Qualcomm AI Engine Direct - Preserve partition metadata on single-output delegate getitem nodes #20805

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
DannyYuyang-quic wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Qualcomm AI Engine Direct - Preserve partition metadata on single-output delegate getitem nodes #20805

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
5 changes: 3 additions & 2 deletions exir/lowered_backend_module.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -884,10 +884,11 @@ def create_submodule_from_nodes(
# all uses with a getitem call to the 0th index of the result
with gm.graph.inserting_after(submodule_node):
proxy_out = torch.fx.Proxy(submodule_node)[0].node # type: ignore[index]
submodule_node.replace_all_uses_with(proxy_out)
proxy_out.meta["val"] = submodule_node.meta["val"]
submodule_node.replace_all_uses_with(proxy_out, propagate_meta=True)
# Reset the args since it was overwritten in the previous line
proxy_out.args = (submodule_node, 0)
proxy_out.meta.pop("nn_module_stack", None)
proxy_out.meta.pop("source_fn_stack", None)
else:
# fuse_as_graphmodule will automatically propagate the metadata of the
# partition's last node to the getitem nodes that appear after the
Expand Down
Loading