Skip to content
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

Libnd4j: broadcastable pairwise op gradient failing for edge case #6037

Closed
AlexDBlack opened this issue Aug 1, 2018 · 4 comments

Comments

@AlexDBlack
Copy link
Contributor

commented Aug 1, 2018

Passing for most cases, but failing for the following case:
Input 1 shape: [1,6]
Input 2 shape: [3,4,5,1]
Output shape: [3,4,5,6]
Failing ops: sub, div, rsub, rdiv

Test case in Java: (note all prior cases are passing)
https://github.com/deeplearning4j/deeplearning4j/blob/0c7d15e2cc3d2167672f6ea96dbe45dec92c1f0e/nd4j/nd4j-backends/nd4j-tests/src/test/java/org/nd4j/autodiff/opvalidation/MiscOpValidation.java#L258

https://gist.github.com/AlexDBlack/774d6a86757d0a246a28789bdd7da471

@AlexDBlack AlexDBlack added C++ Bug labels Aug 1, 2018

@shugeo shugeo self-assigned this Aug 10, 2018

@shugeo

This comment has been minimized.

Copy link
Contributor

commented Aug 10, 2018

there are subtract, reversesubtract, divide and reversedivide broadcastable ops.
The problem reproduces only with reversesubtract and reversedivide.

@AlexDBlack

This comment has been minimized.

Copy link
Contributor Author

commented Aug 16, 2018

Looking at the commit history - is this fixed by #6134 ?
Or is there more to do here?

@raver119

This comment has been minimized.

Copy link
Contributor

commented Aug 16, 2018

Hypothetically should be fixed, at least c++ tests pass now.
Please confirm fix on your end.

@shugeo shugeo closed this Aug 21, 2018

@lock

This comment has been minimized.

Copy link

commented Sep 21, 2018

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.

@lock lock bot locked and limited conversation to collaborators Sep 21, 2018

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
3 participants
You can’t perform that action at this time.