-
Notifications
You must be signed in to change notification settings - Fork 84
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.
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
Properly test non-contiguous input tensors. #34
Conversation
Co-authored-by: Ivan Yashchuk <IvanYashchuk@users.noreply.github.com>
@t-vi this is ready to merge. The CI failure is unrelated and happens at head too. |
CI is passing now. This PR is a port of https://github.com/Lightning-AI/lit-thunder-LEGACY/pull/2478, which was already approved. So it's ready to merge. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Seems good, thank you @wujingyue @IvanYashchuk @vedaanta-nvidia
The current code generates input tensors of wrong shapes, [B,H,D,S] instead of [B,H,S,D]. These wrong shapes unintentionally bypass this check, so this set of input tensors are not fed into cudnnex at all.