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 2020/08/07 19:00:51 UTC
[GitHub] [incubator-mxnet] sxjscience opened a new pull request #18876: [Numpy] Add hybridization test to loss layers
sxjscience opened a new pull request #18876:
URL: https://github.com/apache/incubator-mxnet/pull/18876
## Description ##
Small fix to the loss layer + add hybridization tests.
Since the `exclude` flag is not supported yet, it will raise an error if `batch_axis` is not 0.
----------------------------------------------------------------
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] sxjscience commented on a change in pull request #18876: [Numpy][Bugfix] Add hybridization test to loss layers
Posted by GitBox <gi...@apache.org>.
sxjscience commented on a change in pull request #18876:
URL: https://github.com/apache/incubator-mxnet/pull/18876#discussion_r467322411
##########
File path: python/mxnet/symbol/numpy/_symbol.py
##########
@@ -8012,7 +8012,7 @@ def diagonal(a, offset=0, axis1=0, axis2=1):
# pylint:disable=redefined-outer-name, too-many-arguments
@set_module('mxnet.symbol.numpy')
-def sum(a, axis=None, dtype=None, out=None, keepdims=None, initial=None, where=None):
+def sum(a, axis=None, dtype=None, out=None, keepdims=False, initial=None, where=None):
Review comment:
@szha Figured out a bug in `symbol.sum` which does not hybridization. Minimal reproducible:
```python
import mxnet as mx
from mxnet.gluon.nn import HybridBlock
mx.npx.set_np()
class Foo(HybridBlock):
def hybrid_forward(self, F, x):
return F.np.sum(x)
foo = Foo()
foo.hybridize()
out = foo(mx.np.ones((10, 10)))
out.asnumpy()
```
----------------------------------------------------------------
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] sxjscience merged pull request #18876: [Numpy][Bugfix] Add hybridization test to loss layers
Posted by GitBox <gi...@apache.org>.
sxjscience merged pull request #18876:
URL: https://github.com/apache/incubator-mxnet/pull/18876
----------------------------------------------------------------
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 #18876: [Numpy] Add hybridization test to loss layers
Posted by GitBox <gi...@apache.org>.
mxnet-bot commented on pull request #18876:
URL: https://github.com/apache/incubator-mxnet/pull/18876#issuecomment-670663562
Hey @sxjscience , 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**: [windows-cpu, website, clang, sanity, edge, unix-cpu, centos-cpu, unix-gpu, centos-gpu, miscellaneous, windows-gpu]
***
_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] sxjscience commented on pull request #18876: [Numpy][Bugfix] Add hybridization test to loss layers
Posted by GitBox <gi...@apache.org>.
sxjscience commented on pull request #18876:
URL: https://github.com/apache/incubator-mxnet/pull/18876#issuecomment-670779210
Figured out that the all reduce operators do not support keepdims=None:
Root cause: https://github.com/apache/incubator-mxnet/blob/dde635f8c7666e205816da8ca4e519d0d213523f/src/operator/numpy/np_broadcast_reduce_op.h#L43
----------------------------------------------------------------
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