Skip to content

Fix conversion of ONNX Clip nodes when optional inputs are missing. - #80

Open
asnodin wants to merge 2 commits into
Talmaj:masterfrom
asnodin:asnodin/convert_clip_fix
Open

Fix conversion of ONNX Clip nodes when optional inputs are missing.#80
asnodin wants to merge 2 commits into
Talmaj:masterfrom
asnodin:asnodin/convert_clip_fix

Conversation

@asnodin

@asnodin asnodin commented Jul 23, 2026

Copy link
Copy Markdown

ONNX represents omitted optional node inputs as empty strings. These were being converted into PyTorch arguments that yield incorrect output shapes for models containing Clip nodes without optional min or max.

Added a regression test containing a Clip node with an omitted max input (that is now correctly ignored during conversion).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant