eric-haibin-lin commented on issue #9811: How to reinitialize the begin_state
of an RNN
URL:
https://github.com/apache/incubator-mxnet/issues/9811#issuecomment-366415314
(If you're using the module API) You can explicitly set the begin state via
the `module.set_states` API if you declare
DickJC123 commented on a change in pull request #9791: CI test randomness 3
URL: https://github.com/apache/incubator-mxnet/pull/9791#discussion_r168910658
##
File path: tests/python/unittest/test_operator.py
##
@@ -4401,7 +4522,10 @@ def test_stack():
DickJC123 opened a new issue #9816: Dropout may mask values even when ratio=0.0
URL: https://github.com/apache/incubator-mxnet/issues/9816
One reasonable expectation is that a dropout layout would pass no values
when the dropout ratio=1.0 and would pass all values with the dropout
juliusshufan commented on issue #9805: Enable the reporting of cross-entropy or
nll loss value when training CNN network using the models defined by
example/image-classification
URL: https://github.com/apache/incubator-mxnet/pull/9805#issuecomment-366410556
@piiswrong @cjolivier01 May I
juliusshufan commented on issue #9805: Enable the reporting of cross-entropy or
nll loss value when training CNN network using the models defined by
example/image-classification
URL: https://github.com/apache/incubator-mxnet/pull/9805#issuecomment-366410556
@piiswrong @cjolivier01 May I
juliusshufan commented on issue #9805: Enable the reporting of cross-entropy or
nll loss value when training CNN network using the models defined by
example/image-classification
URL: https://github.com/apache/incubator-mxnet/pull/9805#issuecomment-366410556
@piiswrong @cjolivier01 May I
pengzhao-intel commented on a change in pull request #9810: remove
MKL_EXPERIMENTAL and update make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#discussion_r168908639
##
File path: MKL_README.md
##
@@ -17,46 +17,3 @@ Installing and enabling
pengzhao-intel commented on issue #9810: remove MKL_EXPERIMENTAL and update
make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#issuecomment-366407186
@ashokei there's a test for MKLML installation,
tests/python/cpu/test_mklml.py.
Is it possible to update
jeremiedb commented on issue #9358: Why does running 1 round of an MXNET model
training produce Train-mse=NaN?
URL:
https://github.com/apache/incubator-mxnet/issues/9358#issuecomment-366400747
This looks like a bug. There should be a reset of the iterator prior to the
forward-backward
DickJC123 commented on a change in pull request #9791: CI test randomness 3
URL: https://github.com/apache/incubator-mxnet/pull/9791#discussion_r168904307
##
File path: tests/python/unittest/test_operator.py
##
@@ -4401,7 +4522,10 @@ def test_stack():
marcoabreu commented on issue #9809: register optimizers only once in
CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-366393203
@larroy could you review?
This is an automated message from
marcoabreu commented on issue #9809: register optimizers only once in
CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-366393203
@larroy could you verify?
This is an automated message from
marcoabreu commented on a change in pull request #9791: CI test randomness 3
URL: https://github.com/apache/incubator-mxnet/pull/9791#discussion_r168898453
##
File path: tests/python/unittest/test_operator.py
##
@@ -4401,7 +4522,10 @@ def test_stack():
ashokei commented on a change in pull request #9810: remove MKL_EXPERIMENTAL
and update make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#discussion_r168897511
##
File path: MKL_README.md
##
@@ -17,46 +17,3 @@ Installing and enabling the full
reminisce commented on issue #9802: No support for slicing arrays along
multiple axes
URL:
https://github.com/apache/incubator-mxnet/issues/9802#issuecomment-366390528
I think the op `slice` should suffice. Can you give an example where `slice`
cannot do multiple-axis slicing?
marcoabreu commented on a change in pull request #9810: remove MKL_EXPERIMENTAL
and update make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#discussion_r168897046
##
File path: docker_multiarch/arm.crosscompile.android.mk
##
@@ -82,21 +82,6
marcoabreu commented on a change in pull request #9810: remove MKL_EXPERIMENTAL
and update make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#discussion_r168896988
##
File path: MKL_README.md
##
@@ -17,46 +17,3 @@ Installing and enabling the
zhanghang1989 commented on a change in pull request #9738: Add Support for int64
URL: https://github.com/apache/incubator-mxnet/pull/9738#discussion_r168894798
##
File path: src/operator/mxnet_op.h
##
@@ -204,6 +204,15 @@ inline int get_num_threads(const int N) {
}
zhanghang1989 commented on issue #9688: Adapt operators from PyTorch, will keep
adding
URL: https://github.com/apache/incubator-mxnet/pull/9688#issuecomment-366386448
Agree. I will make the changes.
This is an automated
ashokei commented on a change in pull request #9810: remove MKL_EXPERIMENTAL
and update make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#discussion_r168891508
##
File path: MKL_README.md
##
@@ -1,62 +0,0 @@
-# Full MKL Installation
-
-##
ashokei commented on a change in pull request #9810: remove MKL_EXPERIMENTAL
and update make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#discussion_r168890166
##
File path: docs/faq/perf.md
##
@@ -18,10 +18,7 @@ Performance is mainly
Ishitori opened a new issue #9814: C++ Convolution layer usage produces
"Segmentation fault"
URL: https://github.com/apache/incubator-mxnet/issues/9814
## Description
Compiled version 1.1.0 C++ package with `make -j USE_OPENCV=0
USE_CPP_PACKAGE=1` on Mac and trying to run LeNet
zheng-da commented on a change in pull request #9810: remove MKL_EXPERIMENTAL
and update make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#discussion_r168885870
##
File path: docs/faq/perf.md
##
@@ -18,10 +18,7 @@ Performance is mainly
zheng-da commented on a change in pull request #9810: remove MKL_EXPERIMENTAL
and update make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#discussion_r168885655
##
File path: MKL_README.md
##
@@ -1,62 +0,0 @@
-# Full MKL Installation
-
-##
szha commented on issue #9358: Why does running 1 round of an MXNET model
training produce Train-mse=NaN?
URL:
https://github.com/apache/incubator-mxnet/issues/9358#issuecomment-366373020
@jeremiedb would you help on this?
szha commented on issue #9358: Why does running 1 round of an MXNET model
training produce Train-mse=NaN?
URL:
https://github.com/apache/incubator-mxnet/issues/9358#issuecomment-366372523
@jeremydb would you help on this?
szha commented on issue #9358: Why does running 1 round of an MXNET model
training produce Train-mse=NaN?
URL:
https://github.com/apache/incubator-mxnet/issues/9358#issuecomment-366372523
@jeremydb would you help on this?
ashokei commented on issue #9810: remove MKL_EXPERIMENTAL and update make files
for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810#issuecomment-366372304
@zheng-da @piiswrong can you please review ... these are minor changes to
remove old references to MKL_EXPERIMENTAL
indhub commented on issue #8639: [WIP] Gluon object detection
URL: https://github.com/apache/incubator-mxnet/pull/8639#issuecomment-366361216
Which separate repo was this moved to?
This is an automated message from the Apache
samhodge opened a new issue #9813: Unable to save gluon model to symbolic
network : neural style
URL: https://github.com/apache/incubator-mxnet/issues/9813
## Description
see: https://github.com/zhanghang1989/MXNet-Gluon-Style-Transfer/issues/2
within
DickJC123 opened a new issue #9812: Unrepeatable
test_gluon_model_zoo_gpu.py:test_training CI failures seen.
URL: https://github.com/apache/incubator-mxnet/issues/9812
During the development of the ci_test_randomness3 #9791 PR, failures were
seen in
ArmageddonKnight opened a new issue #9811: How to reinitialize the begin_state
of an RNN
URL: https://github.com/apache/incubator-mxnet/issues/9811
I now have an RNN, of which the initial hidden states are initialized to all
zeros:
```
rnn = mx.rnn.SequentialRNNCell()
chsin commented on issue #9809: register optimizers only once in CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-366324817
@marcoabreu, OK, where are the instructions for adding a unit test to the
cpp package? I see that in
marcoabreu commented on issue #9809: register optimizers only once in
CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-366328371
Oh you're right, we don't have a structure to test cpp package properly. My
bad...
But yes, everything in cpp-package
chsin commented on issue #9809: register optimizers only once in CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-366324817
@marcoabreu, OK, where are the instructions for adding a test to the cpp
package? I see that in
chsin commented on issue #9809: register optimizers only once in CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-366324817
@marcoabreu, OK, where are the instructions for adding a test to the cpp
package? I see that in
chsin commented on issue #9809: register optimizers only once in CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-366324817
@marcoabreu, OK, where are the instructions for adding a test to the cpp
package?
ashokei opened a new pull request #9810: remove MKL_EXPERIMENTAL and update
make files for MKL-DNN
URL: https://github.com/apache/incubator-mxnet/pull/9810
## Description ##
remove MKL_EXPERIMENTAL and update make files for MKL-DNN
## Checklist ##
### Essentials ###
-
marcoabreu commented on issue #9809: register optimizers only once in
CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-366294648
Could you please add a test for this case?
This is an
chsin opened a new pull request #9809: register optimizers only once
URL: https://github.com/apache/incubator-mxnet/pull/9809
## Description ##
Regards issue #9800. The bug was that CPP-Package won't allow more than one
optimizer to be created. See comments for more details.
##
marcoabreu commented on issue #9807: temporary solution for instancenorm, will
refactor using backend
URL: https://github.com/apache/incubator-mxnet/pull/9807#issuecomment-366271611
Okay, thanks for elaborating
This is an
szha commented on issue #9807: temporary solution for instancenorm, will
refactor using backend
URL: https://github.com/apache/incubator-mxnet/pull/9807#issuecomment-366266067
The API was only recently added in master and hasn?t been introduced in
release yet.
larroy opened a new pull request #9808: Ssd camera demo
URL: https://github.com/apache/incubator-mxnet/pull/9808
## Description ##
This uses a webcam or video capture device to run SSD in real time.
## Checklist ##
### Essentials ###
- [ ] Passed code style checking
sbodenstein commented on issue #9543: Variable Length Support for cuDNN RNN
URL:
https://github.com/apache/incubator-mxnet/issues/9543#issuecomment-366230527
Any comments on this?
This is an automated message from the Apache
marcoabreu commented on issue #9807: temporary solution for instancenorm, will
refactor using backend
URL: https://github.com/apache/incubator-mxnet/pull/9807#issuecomment-366226037
Correct me if I'm wrong, but isn't this an API change?
zhanghang1989 opened a new pull request #9807: temporary solution for
instancenorm, will refactor using backend
URL: https://github.com/apache/incubator-mxnet/pull/9807
## Description ##
(Brief description on what this PR is about)
## Checklist ##
### Essentials ###
- [ ]
marcoabreu opened a new issue #9806: Some build failures on Windows don't stop
Jenkins jobs
URL: https://github.com/apache/incubator-mxnet/issues/9806
Some build failures like in
http://jenkins.mxnet-ci.amazon-ml.com/blue/organizations/jenkins/incubator-mxnet/detail/PR-9798/1/pipeline/432
marcoabreu commented on issue #9799: Cleaned up image classification cpp example
URL: https://github.com/apache/incubator-mxnet/pull/9799#issuecomment-366206805
Could you please also test Windows?
This is an automated message
marcoabreu commented on issue #9803: R Metrics
URL: https://github.com/apache/incubator-mxnet/pull/9803#issuecomment-366205820
Hi, could you provide your benchmark values for this change?
This is an automated message from the
marcoabreu commented on issue #9791: CI test randomness 3
URL: https://github.com/apache/incubator-mxnet/pull/9791#issuecomment-366204903
I'd like to see this feature and the stated assumptions being tested, but
otherwise it LGTM.
marcoabreu commented on a change in pull request #9791: CI test randomness 3
URL: https://github.com/apache/incubator-mxnet/pull/9791#discussion_r168726380
##
File path: tests/python/unittest/test_operator.py
##
@@ -2265,18 +2317,18 @@ def
marcoabreu commented on a change in pull request #9791: CI test randomness 3
URL: https://github.com/apache/incubator-mxnet/pull/9791#discussion_r168726176
##
File path: tests/python/unittest/test_gluon.py
##
@@ -530,13 +562,19 @@ def __init__(self, **kwargs):
marcoabreu commented on a change in pull request #9791: CI test randomness 3
URL: https://github.com/apache/incubator-mxnet/pull/9791#discussion_r168725979
##
File path: tests/python/unittest/common.py
##
@@ -30,3 +35,190 @@ def assertRaises(expected_exception, func,
lebeg commented on a change in pull request #9799: Cleaned up image
classification cpp example
URL: https://github.com/apache/incubator-mxnet/pull/9799#discussion_r168725721
##
File path:
example/image-classification/predict-cpp/image-classification-predict.cc
##
@@
lebeg commented on issue #9799: Cleaned up image classification cpp example
URL: https://github.com/apache/incubator-mxnet/pull/9799#issuecomment-366203141
Listed tested platforms in the description
This is an automated
lebeg commented on a change in pull request #9799: Cleaned up image
classification cpp example
URL: https://github.com/apache/incubator-mxnet/pull/9799#discussion_r168723620
##
File path: CMakeLists.txt
##
@@ -671,8 +672,7 @@ if(USE_CPP_PACKAGE)
juliusshufan closed pull request #9793: Enable the reporting of cross-entropy
or nll loss value when training CNN network using the models defined by
example/image-classification
URL: https://github.com/apache/incubator-mxnet/pull/9793
juliusshufan commented on issue #9793: Enable the reporting of cross-entropy or
nll loss value when training CNN network using the models defined by
example/image-classification
URL: https://github.com/apache/incubator-mxnet/pull/9793#issuecomment-366183054
As my repo is out-of-date to
juliusshufan commented on a change in pull request #9793: Enable the reporting
of cross-entropy or nll loss value when training CNN network using the models
defined by example/image-classification
URL: https://github.com/apache/incubator-mxnet/pull/9793#discussion_r168705538
##
juliusshufan opened a new pull request #9805: Enable the reporting of
cross-entropy or nll loss value when training CNN network using the models
defined by example/image-classification
URL: https://github.com/apache/incubator-mxnet/pull/9805
## Description ##
MxNET already implements
60 matches
Mail list logo