Skip to content

Implement grid_sampler in aten#8929

Closed
li-roy wants to merge 4 commits intopytorch:masterfrom
li-roy:nativegrid
Closed

Implement grid_sampler in aten#8929
li-roy wants to merge 4 commits intopytorch:masterfrom
li-roy:nativegrid

Conversation

@li-roy
Copy link
Contributor

@li-roy li-roy commented Jun 27, 2018

Partially addresses #8928.

Maybe #7273?

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

Copy link
Contributor

@zou3519 zou3519 left a comment

Choose a reason for hiding this comment

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

Looks good! I left some questions. The one point of action would be to add a test for the (input.size(1) > 1024 and dim == 4) case

This comment was marked as off-topic.

This comment was marked as off-topic.

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.

@li-roy has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

@ezyang
Copy link
Contributor

ezyang commented Jun 27, 2018

@pytorchbot retest this please

@ezyang
Copy link
Contributor

ezyang commented Jun 28, 2018

22:40:25 FAIL: test_grid_sample (__main__.TestNN)
22:40:25 ----------------------------------------------------------------------
22:40:25 Traceback (most recent call last):
22:40:25   File "test_nn.py", line 4708, in test_grid_sample
22:40:25     test_cpu_against_cuda(N, C, H, W, padding_mode)
22:40:25   File "test_nn.py", line 4646, in test_cpu_against_cuda
22:40:25     test_shape(N, C, H, W, H, W, padding_mode)
22:40:25   File "test_nn.py", line 4632, in test_shape
22:40:25     self.assertEqual(grid_cpu.grad, grid_cuda.grad, prec=5e-5)
22:40:25   File "/var/lib/jenkins/workspace/test/common.py", line 318, in assertEqual
22:40:25     assertTensorsEqual(x, y)
22:40:25   File "/var/lib/jenkins/workspace/test/common.py", line 310, in assertTensorsEqual
22:40:25     self.assertLessEqual(max_err, prec, message)
22:40:25 AssertionError: tensor(41693651666285782607020689418557447054364822151646846543972235746272785307628883019700693160086740858065237563435119047359851402875793184184988497299293337780641492578057095198773151115203507315583266163122786152348900255934909802315279944257991753435893013887381159690437409208881152786432.) not less than or equal to 5e-05 : 

@zou3519
Copy link
Contributor

zou3519 commented Jun 28, 2018

That is a very big number

@li-roy
Copy link
Contributor Author

li-roy commented Jul 7, 2018

upon further testing, this just doesn't work on any cudnn version. it throws error on some and gives wrong output for others.

Copy link
Contributor

@zou3519 zou3519 left a comment

Choose a reason for hiding this comment

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

LGTM

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.

@li-roy has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

zdevito pushed a commit to zdevito/ATen that referenced this pull request Jul 10, 2018
Summary:
Partially addresses #8928.

Maybe #7273?
Pull Request resolved: pytorch/pytorch#8929

Reviewed By: ezyang

Differential Revision: D8668919

Pulled By: li-roy

fbshipit-source-id: 8ad07b224d2ab211c274c4c10f042501efaae32c
zdevito pushed a commit to zdevito/ATen that referenced this pull request Jul 13, 2018
Summary:
Partially addresses #8928.

Maybe #7273?
Pull Request resolved: pytorch/pytorch#8929

Reviewed By: ezyang

Differential Revision: D8668919

Pulled By: li-roy

fbshipit-source-id: 8ad07b224d2ab211c274c4c10f042501efaae32c
goodlux pushed a commit to goodlux/pytorch that referenced this pull request Aug 15, 2018
Summary:
Partially addresses pytorch#8928.

Maybe pytorch#7273?
Pull Request resolved: pytorch#8929

Reviewed By: ezyang

Differential Revision: D8668919

Pulled By: li-roy

fbshipit-source-id: 8ad07b224d2ab211c274c4c10f042501efaae32c
@ezyang ezyang added the merged label Jun 26, 2019
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.

6 participants