You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by GitBox <gi...@apache.org> on 2020/07/14 11:53:26 UTC

[GitHub] [incubator-mxnet] bgawrych opened a new pull request #18711: Fix (log_)softmax backward on empty ndarray

bgawrych opened a new pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711


   ## Description ##
   This pull request fixes #18710 issue. Fixed in the same way as forward pass
   
   ## Checklist ##
   ### Essentials ###
   - [x] Changes are complete (i.e. I finished coding on this PR)
   - [x] All changes have test coverage:
   - Unit tests are added for small changes to verify correctness (e.g. adding a new operator)
   - [x] To the best of my knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change
   
   


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] bgawrych commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
bgawrych commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-658576492


   > Could you please add two unit-tests `softmax and log_softmax with empty ndarray` into `tests/python/unittest/test_operator.py` ? Check their backward procedure when input is empty.
   > 
   > Thank you!
   
   @wkcn  Please correct me if I'm wrong but doesn't empty array (e.g. [3,0,4]) is allowed only when numpy semantics is turned on and legacy mode doesn't support this? 


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] DickJC123 edited a comment on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
DickJC123 edited a comment on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-659516305


   My PR is having the same 'worker crash' problem this PR is having.  Without understanding the root cause, I tried marking the failing test in unittest/test_sparse_operator.py as serial, and I got no similar failures.  So you might give this a try:
   ```
   @with_seed()
   @pytest.mark.serial
   def test_elemwise_binary_ops():
   ```


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] wkcn commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
wkcn commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-658594952


   @bgawrych sorry for my mistake. We does not need to add test for legacy operator.
   You can add test into tests/python/unittest/test_numpy_op.py


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-658649633


   Jenkins CI successfully triggered : [centos-cpu]


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] bgawrych commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
bgawrych commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-658649053


   @wkcn It's already done :) Function test_npx_softmax (tests/python/unittest/test_numpy_op.py:1916) is checking empty forward and backward pass
   


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] DickJC123 commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
DickJC123 commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-659119022


   I'm working on a PR that is mostly failing due to the issues this PR fixes, so I'll be happy to see this merged.  Are the CI failures you're seeing related to this commit, thus requiring a follow-up commit, or are you done?  I'm thinking about cherry-picking you're present commit onto my PR so I can get to a cleaner CI.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] wkcn commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
wkcn commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-658726394


   @mxnet-bot run ci [centos-cpu]


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] DickJC123 edited a comment on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
DickJC123 edited a comment on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-659119022


   I'm working on a PR that is mostly failing due to the issues this PR fixes, so I'll be happy to see this merged.  Are the CI failures you're seeing related to this commit, thus requiring a follow-up commit, or are you done?  I'm thinking about cherry-picking your present commit onto my PR so I can get to a cleaner CI.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-658726455


   Jenkins CI successfully triggered : [centos-cpu]


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] wkcn commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
wkcn commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-658531787


   Could you please add two unit-tests `softmax and log_softmax with empty ndarray` into `tests/python/unittest/test_operator.py` ? Check their backward procedure when input is empty.
   
   Thank you!


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] bgawrych commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
bgawrych commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-658649603


   @mxnet-bot run ci [centos-cpu]


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] bgawrych commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
bgawrych commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-660915356


   Closing, as fix have been merged to master in #18694


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] bgawrych closed pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
bgawrych closed pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711


   


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-658136852


   Hey @bgawrych , Thanks for submitting the PR 
   All tests are already queued to run once. If tests fail, you can trigger one or more tests again with the following commands: 
   - To trigger all jobs: @mxnet-bot run ci [all] 
   - To trigger specific jobs: @mxnet-bot run ci [job1, job2] 
   *** 
   **CI supported jobs**: [unix-gpu, clang, unix-cpu, windows-gpu, centos-gpu, website, windows-cpu, miscellaneous, edge, centos-cpu, sanity]
   *** 
   _Note_: 
    Only following 3 categories can trigger CI :PR Author, MXNet Committer, Jenkins Admin. 
   All CI tests must pass before the PR can be merged. 
   


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] DickJC123 commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
DickJC123 commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-659516305


   My PR is having the same 'worker crash' problem this PR is having.  Without understanding the root cause, I tried marking the failing test in unittest/test_sparse_operator.py as serial, and I got no similar failures.  So you might give this a try:
   ```
   @with_seed()
   @pytest.mark.serial
    def test_elemwise_binary_ops():
   ```


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] mxnet-bot commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-659123531


   Jenkins CI successfully triggered : [centos-cpu]


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [incubator-mxnet] szha commented on pull request #18711: Fix (log_)softmax backward on empty ndarray

Posted by GitBox <gi...@apache.org>.
szha commented on pull request #18711:
URL: https://github.com/apache/incubator-mxnet/pull/18711#issuecomment-659123504


   The errors seem unrelated. @mxnet-bot run ci [centos-cpu]


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org