Skip to content

Do not try to fold op.SplitToSequence when split is None #2550

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
titaiwangms merged 1 commit into from
Sep 8, 2025
Merged

Do not try to fold op.SplitToSequence when split is None #2550

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: 3 additions & 0 deletions onnxscript/optimizer/_constant_folding.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -784,6 +784,9 @@ def split_to_sequence(node: ir.Node, op, state: OptimizerState) -> ReturnValue:
This allows downstream `SequenceAt` users to be replaced by `split_x` accordingly.
"""
input = node.inputs[0]
if len(node.inputs) == 1:
# split is not provided
Copy link
Copy Markdown
Collaborator

@justinchuby justinchuby Sep 9, 2025

Choose a reason for hiding this comment

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

When split is not provided it is a scalar 1. Can we use this info?

Copy link
Copy Markdown
Contributor Author

@titaiwangms titaiwangms Sep 9, 2025 •
edited
Loading

Choose a reason for hiding this comment

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

Is it? The issue @xadupre brought up was only one input. model id: facebook/sam-vit-base, so it hit indexerror.

Copy link
Copy Markdown
Collaborator

@justinchuby justinchuby Sep 9, 2025

Choose a reason for hiding this comment

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

I think at least according to the spec: https://onnx.ai/onnx/operators/onnx__SplitToSequence.html#summary

Copy link
Copy Markdown
Contributor Author

@titaiwangms titaiwangms Sep 9, 2025

Choose a reason for hiding this comment

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

onnxscript/onnxscript/function_libs/torch_lib/ops/core.py

Line 8732 in cec5396

return op.SplitToSequence(self, axis=dim, keepdims=False)

Copy link
Copy Markdown
Contributor Author

@titaiwangms titaiwangms Sep 9, 2025

Choose a reason for hiding this comment

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

Probably came from here

return None
split = node.inputs[1]
output = node.outputs[0]

Expand Down
Loading
⚡