Skip to content

Conversation

@neginraoof
Copy link
Contributor

This would also improve error handling for interpolate with 'area' mode.

@dr-ci
Copy link

dr-ci bot commented Aug 14, 2020

💊 CI failures summary and remediations

As of commit c8fc4bf (more details on the Dr. CI page):


Commit c8fc4bf was recently pushed. Waiting for builds...


This comment was automatically generated by Dr. CI (expand for details).Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions on the GitHub issue tracker or post in the (internal) Dr. CI Users group.

See how this bot performed.

This comment has been revised 52 times.

@mrshenli mrshenli added module: onnx Related to torch.onnx triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module labels Aug 18, 2020
@mrshenli mrshenli requested a review from houseroad August 18, 2020 17:31
Copy link
Collaborator

@BowenBao BowenBao left a comment

Choose a reason for hiding this comment

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

LG, thanks

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.

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

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.

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

@neginraoof
Copy link
Contributor Author

@bzinodev
The failure in Jenkins: pr/pytorch-linux-bionic-rocm3.7-py3.6
is unrelated. I see this failure on other recent PR's as well.

Would you please let me know if this currently blocking merge? Or is there anything I can do to unblock this?

@codecov
Copy link

codecov bot commented Sep 1, 2020

Codecov Report

Merging #43032 into master will decrease coverage by 0.00%.
The diff coverage is 0.00%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master   #43032      +/-   ##
==========================================
- Coverage   68.05%   68.04%   -0.01%     
==========================================
  Files         396      396              
  Lines       51235    51238       +3     
==========================================
  Hits        34867    34867              
- Misses      16368    16371       +3     
Impacted Files Coverage Δ
torch/onnx/symbolic_opset9.py 34.66% <0.00%> (-0.13%) ⬇️
torch/onnx/utils.py 72.06% <ø> (ø)
torch/testing/_internal/expecttest.py 78.57% <0.00%> (+1.02%) ⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 993628c...c8fc4bf. Read the comment docs.

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.

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

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.

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

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.

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

@neginraoof
Copy link
Contributor Author

@bzinodev
Can we please merge this PR?
Thanks!

@neginraoof neginraoof changed the title [ONNX] Imropve error handling for adaptive_pool [ONNX] Improve error handling for adaptive_pool Sep 17, 2020
@neginraoof
Copy link
Contributor Author

@bzinodev
Can we please merge this PR?
Thanks!

@gchanan gchanan added this to the 1.7.0 milestone Sep 21, 2020
@neginraoof
Copy link
Contributor Author

@bzinodev Can we merge this please? Thanks!

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.

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

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.

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

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.

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

@neginraoof
Copy link
Contributor Author

cc @bzinodev
This PR is rebased and ready for merge. Thanks a lot.

if mod != [0] * len(mod):
if output_size == [1] * len(output_size):
return g.op("GlobalMaxPool", input), None
return _unimplemented(name, 'output size that are not factor of input size')
Copy link
Contributor

Choose a reason for hiding this comment

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

Old behavior a warning is emitted. It is changed to a runtime error? there is one internal test that fails. Is this something we want?

Copy link
Contributor

Choose a reason for hiding this comment

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

Note other ops all uses _unimplemented

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Before my update, this line in the symbolic conversion was throwing a warning instead of an error. Hence, the ATen node in the graph was not converted to ONNX, and was left as ATen.
By default, this behavior is not accepted. But exporter has a configuration mode that accepts combination of ONNX and ATen nodes. It's called ONNX_ATEN_FALLBACK.
It could be that the internal test case you see is failing is exporting with the ONNX_ATEN_FALLBACK mode.
I'll send a fix.

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.

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

@neginraoof
Copy link
Contributor Author

@bzinodev Could you please merge the last commit here? There was a typo that was fixed after I addressed your comment.

@dzhulgakov
Copy link
Collaborator

Hey, this PR unfortunately broke master (probably the missing typo you mentioned). I'm unlanding it, please resubmit the new PR with the fix, @bzinodev or myself can help you land it

@facebook-github-bot
Copy link
Contributor

@bzinodev merged this pull request in 45ddeb5.

@neginraoof
Copy link
Contributor Author

@bzinodev @dzhulgakov The new duplicate PR is: #45874
Could you please help merge this new one?
Thanks!

facebook-github-bot pushed a commit that referenced this pull request Oct 7, 2020
Summary:
Duplicate of #43032
This update would also improve error handling for interpolate with 'area' mode.

Pull Request resolved: #45874

Reviewed By: albanD

Differential Revision: D24141266

Pulled By: bzinodev

fbshipit-source-id: 7559f1d6af4f1ef3507c15a1aee76fe01fa433cd
neginraoof added a commit to neginraoof/pytorch that referenced this pull request Oct 9, 2020
Summary:
Duplicate of pytorch#43032
This update would also improve error handling for interpolate with 'area' mode.

Pull Request resolved: pytorch#45874

Reviewed By: albanD

Differential Revision: D24141266

Pulled By: bzinodev

fbshipit-source-id: 7559f1d6af4f1ef3507c15a1aee76fe01fa433cd
malfet pushed a commit that referenced this pull request Oct 12, 2020
Summary:
Duplicate of #43032
This update would also improve error handling for interpolate with 'area' mode.

Pull Request resolved: #45874

Reviewed By: albanD

Differential Revision: D24141266

Pulled By: bzinodev

fbshipit-source-id: 7559f1d6af4f1ef3507c15a1aee76fe01fa433cd
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Merged module: onnx Related to torch.onnx open source triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module

Projects

None yet

Development

Successfully merging this pull request may close these issues.

9 participants