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

Shyrma deconv3d #6387

Merged
merged 10 commits into from Sep 10, 2018

Conversation

@shyrma
Copy link
Contributor

commented Sep 6, 2018

  • implementation and testing of deconv3d op
  • correct input validation in strided_slice

@shyrma shyrma requested a review from raver119 Sep 6, 2018

@AlexDBlack

This comment has been minimized.

Copy link
Contributor

commented Sep 7, 2018

If you 're using "-999" as a valid input - it's not, and these values shouldn't be used in c++ at all. I assume you're taking them from these test cases:
https://github.com/deeplearning4j/deeplearning4j/blob/2c32b5ca8b9f6189a07faabd2c7da0f19510675c/nd4j/nd4j-backends/nd4j-tests/src/test/java/org/nd4j/autodiff/opvalidation/ShapeOpValidation.java#L426

The idea with those -999 values is basically to make it obvious in the test that they aren't intended to be used - and hopefully to cause a crash if they are used. They don't mean anything as far as the op implementation is concerned.

Edit: the reason they aren't used is because of the begin/end etc masks.

@raver119 raver119 merged commit 4655e18 into master Sep 10, 2018

0 of 2 checks passed

codeclimate Code Climate encountered an error attempting to analyze this pull request.
Details
continuous-integration/jenkins/pr-head This commit has test failures
Details

@raver119 raver119 deleted the shyrma_deconv3d branch Sep 10, 2018

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