Skip to content

Conversation

@ngimel
Copy link
Collaborator

@ngimel ngimel commented Jan 2, 2019

Fixes #15647. cc @colesbury.

@soumith
Copy link
Contributor

soumith commented Jan 2, 2019

if possible, add a test covering the case.

Copy link
Contributor

Choose a reason for hiding this comment

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

Please also add a comment explaining why do we need to test this. Is it that we use different kernels for large vs small inputs? Those things are super helpful when refactoring the tests.

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

There is now a single kernel for all reductions, with the different options. I'm not sure all the possible options are thoroughly tested, it would be a good idea to test them, but I don't know exactly in which cases which options are triggered.

Copy link
Contributor

@facebook-github-bot facebook-github-bot left a comment

Choose a reason for hiding this comment

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

@soumith is landing this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

@soumith
Copy link
Contributor

soumith commented Jan 3, 2019

fixed via #15647 which is landing

@soumith soumith closed this Jan 3, 2019
facebook-github-bot pushed a commit that referenced this pull request Jan 3, 2019
Summary:
Fixes #15647. cc colesbury.
Pull Request resolved: #15653

Differential Revision: D13571132

Pulled By: soumith

fbshipit-source-id: 8f25943c974b3b931f4528e0e0a370bc095dab51
zdevito pushed a commit to zdevito/ATen that referenced this pull request Jan 3, 2019
Summary:
Fixes #15647. cc colesbury.
Pull Request resolved: pytorch/pytorch#15653

Differential Revision: D13571132

Pulled By: soumith

fbshipit-source-id: 8f25943c974b3b931f4528e0e0a370bc095dab51
@ngimel ngimel deleted the fix_prod branch January 16, 2019 19:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

torch.prod on GPU produces incorrect results on large Tensors

5 participants