asitstands opened a new pull request #9870: New example of custom operator
using RTC
URL: https://github.com/apache/incubator-mxnet/pull/9870
## Description ##
`example/numpy-ops/ndarray_softmax.py` contains a custom softmax
implementation using RTC. However, it extends `NDArrayOp`
anirudh2290 opened a new pull request #9869: Exception handling documentation
URL: https://github.com/apache/incubator-mxnet/pull/9869
## Description ##
Adds documentation for exception handling. Requires #9795 to be merged
This is an automated email from the ASF dual-hosted git repository.
liuyizhi pushed a commit to branch asf-site
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git
commit a402a94223482212a6302d276f19aa15cfe9aacd
Merge: 876167c dde3caf
Author: Yizhi Liu
thinksanky opened a new pull request #52: fixed the release number to 1.1.0
URL: https://github.com/apache/incubator-mxnet-site/pull/52
## Description ##
Updated the release number from 1.0 from 1.1.0
This is an automated
dabraude opened a new issue #9868: MKL and CMake
URL: https://github.com/apache/incubator-mxnet/issues/9868
Looking at the CMake with MKL I noticed there is an option there called
`USE_MKLML_MKL` which has the same docstring as `USE_MKLDNN` ; from what I can
tell it is there for support
juliusshufan opened a new pull request #9867: Using "uniform" Xavier strategy
to initialize the weight for VGG network
URL: https://github.com/apache/incubator-mxnet/pull/9867
## Description ##
This PR provide a potential solution for issue
juliusshufan opened a new issue #9866: The default weight initialization
strategy makes the VGG network difficult to converge when utilizing examples
under 'example/image-classification'
URL: https://github.com/apache/incubator-mxnet/issues/9866
## Description
On trying to train
zheng-da commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367913772
it seems the current modification still can't get rid of all race conditions
in the code. the reason is that we
pengzhao-intel commented on issue #9744: Performance regression when
OMP_NUM_THREADS environment variable is not set
URL:
https://github.com/apache/incubator-mxnet/issues/9744#issuecomment-367911898
The OMP works now from the C code and I think KMP should be.
There's C level API
pengzhao-intel commented on issue #9744: Performance regression when
OMP_NUM_THREADS environment variable is not set
URL:
https://github.com/apache/incubator-mxnet/issues/9744#issuecomment-367911898
There's C level API too, _int kmp_set_affinity_ in OpenMP.
The OMP works now and I
pengzhao-intel commented on issue #9744: Performance regression when
OMP_NUM_THREADS environment variable is not set
URL:
https://github.com/apache/incubator-mxnet/issues/9744#issuecomment-367911898
There're C level API too, _int kmp_set_affinity_ in openmp.
The OMP works now and I
pengzhao-intel commented on issue #9744: Performance regression when
OMP_NUM_THREADS environment variable is not set
URL:
https://github.com/apache/incubator-mxnet/issues/9744#issuecomment-367911898
There're C level API too, _int kmp_set_affinity_ in openmp.
Also in this page:
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-367908298
@marcoabreu can you please review requested Jenkins file changes; This PR is
primarily targeting at removing
zheng-da commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170165614
##
File path: src/ndarray/ndarray.cc
##
@@ -1017,6 +1017,7 @@ inline void
cjolivier01 commented on issue #9744: Performance regression when
OMP_NUM_THREADS environment variable is not set
URL:
https://github.com/apache/incubator-mxnet/issues/9744#issuecomment-367907988
i don?t set the environment variable. I doubt it would
work even if i did, from within the
pengzhao-intel commented on issue #9744: Performance regression when
OMP_NUM_THREADS environment variable is not set
URL:
https://github.com/apache/incubator-mxnet/issues/9744#issuecomment-367904801
Normally, we set KMP_AFFILITY with OMP_NUM_THREADS together.
And when the OMP threads
cjolivier01 commented on a change in pull request #9862: Fix a race condition
in converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170162245
##
File path: src/operator/nn/mkldnn/mkldnn_base.cc
##
@@ -270,9 +270,24 @@
cjolivier01 commented on a change in pull request #9862: Fix a race condition
in converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170161984
##
File path: src/ndarray/ndarray.cc
##
@@ -1017,6 +1017,7 @@ inline void
cjolivier01 commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367902985
if you run the test in a loop (from python) 1000 times, does it fail?
cjolivier01 commented on a change in pull request #9862: Fix a race condition
in converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170161424
##
File path: src/operator/nn/mkldnn/mkldnn_base.cc
##
@@ -270,9 +270,24 @@
feevos commented on issue #9822: gluon HybridBlock wrapper of constant
nd.array, is it possible?
URL:
https://github.com/apache/incubator-mxnet/issues/9822#issuecomment-367901846
Thanks @jmacglashan , based on your suggestions I tried the following
solution:
I am trying to
cjolivier01 commented on issue #9744: Performance regression when
OMP_NUM_THREADS environment variable is not set
URL:
https://github.com/apache/incubator-mxnet/issues/9744#issuecomment-367901652
how to set this programmatically? i?d imagine that by the time this code
executes, the omp
anirudh2290 commented on issue #9813: Unable to save gluon model to symbolic
network : neural style
URL:
https://github.com/apache/incubator-mxnet/issues/9813#issuecomment-367901245
do you have a validation dataset to test the model ?
xiaoheizi123 commented on issue #7162: Cant convert caffe model to mxnet
URL:
https://github.com/apache/incubator-mxnet/issues/7162#issuecomment-367900241
I also cannot convert resnet-50 using **python convert_caffe_modelzoo.py
resnet-50**
`converting scale layer, beta shape =
sxjscience commented on issue #9843: Flaky
test_random.test_exponential_generator @ Python2: MKLDNN-CPU
URL:
https://github.com/apache/incubator-mxnet/issues/9843#issuecomment-367879896
@marcoabreu A wrongly implemented random number generator would typically
have success_rate = 0, i.e,
larroy commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170141588
##
File path: src/ndarray/ndarray.cc
##
@@ -1017,6 +1017,7 @@ inline void
szha closed pull request #9863: Fix broken website build and update version
number
URL: https://github.com/apache/incubator-mxnet/pull/9863
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this
This is an automated email from the ASF dual-hosted git repository.
zhasheng pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/master by this push:
new cea8d2f Fix broken website build,
larroy commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170141588
##
File path: src/ndarray/ndarray.cc
##
@@ -1017,6 +1017,7 @@ inline void
zheng-da commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367876047
It's very difficult to reproduce a race condition in a deterministic way if
it's possible.
larroy commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170139781
##
File path: src/operator/nn/mkldnn/mkldnn_base.cc
##
@@ -270,9 +270,24 @@ void
marcoabreu commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367872237
Would it be possible to create a test that introduces a delay or uses a
for-loop in order to force this race
nswamy closed pull request #9851: Softsign op
URL: https://github.com/apache/incubator-mxnet/pull/9851
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a
This is an automated email from the ASF dual-hosted git repository.
nswamy pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/master by this push:
new 6ebec87 Softsign Activation Function
Maratyszcza commented on issue #9860: [WIP] CMake NNPack support
URL: https://github.com/apache/incubator-mxnet/pull/9860#issuecomment-367873330
@piiswrong There is no bundled repo. The options are to add NNPACK submodule
and let it download all deps at configuration time, or add NNPACK
pengzhao-intel commented on issue #9744: Performance regression when
OMP_NUM_THREADS environment variable is not set
URL:
https://github.com/apache/incubator-mxnet/issues/9744#issuecomment-367873227
Thanks, @cjolivier01, I mean we set this environment in the code as
OMP_NUM_THREADS.
marcoabreu commented on issue #9798: fix cmake
URL: https://github.com/apache/incubator-mxnet/pull/9798#issuecomment-367872877
Would you mind adding it to the build args in the Jenkinsfile?
This is an automated message from
nswamy commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367872369
I am closing this PR. Customer is ok to apply the patch on v.0.11.
This is an
nswamy closed pull request #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull
marcoabreu commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367872237
Would it be possible to create a test that introduces a delay in order to
force this race condition?
yajiedesign commented on issue #9798: fix cmake
URL: https://github.com/apache/incubator-mxnet/pull/9798#issuecomment-367869677
@marcoabreu no,not use in Jenkinsfile.
This is an automated message from the Apache Git Service.
zheng-da commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367869410
The reason I disabled the inference tests because I previously thought the
failure was related to numeric errors
zheng-da commented on issue #9853: Flaky test_operator.test_binary_op
URL:
https://github.com/apache/incubator-mxnet/issues/9853#issuecomment-367868599
I don't have a clue right now. so far we see failures in random generators
and binary operators. it's weird why it fails in these simple
sampathchanda commented on issue #9100: Python - ImportError: cannot import
name 'c_array_buf'
URL:
https://github.com/apache/incubator-mxnet/issues/9100#issuecomment-367866625
Am getting the same issue, even when having mxnet installed from 'pip
install mxnet-cu80==0.11.0'. Any ideas
marcoabreu commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367866215
I'm afraid I still don't really understand. I see that you have re-enabled
this test, but unfortunately I'm
cjolivier01 commented on issue #9853: Flaky test_operator.test_binary_op
URL:
https://github.com/apache/incubator-mxnet/issues/9853#issuecomment-367866040
i don?t know what is invoked in the process of calling test_bmod(). could
be that elemwise_add() isn?t called, or is called before
marcoabreu commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170132857
##
File path: tests/python/gpu/test_gluon_model_zoo_gpu.py
##
@@ -37,8 +37,7
marcoabreu commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170132926
##
File path: tests/python/gpu/test_gluon_model_zoo_gpu.py
##
@@ -99,7 +100,7
marcoabreu commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170132782
##
File path: src/ndarray/ndarray.cc
##
@@ -1017,6 +1017,7 @@ inline void
marcoabreu commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367865250
We don't need fixed seeds since #9791 has been merged. Just make sure that
the with_seed decorator is properly
zheng-da commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170130612
##
File path: src/ndarray/ndarray.cc
##
@@ -1017,6 +1017,7 @@ inline void
zheng-da commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367862145
@marcoabreu enabling the tests can catch the error more easily.
zheng-da commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367861997
@cjolivier01 The seed is set so we know what is the expected result. It's
easier to tell whether CPU or GPU compute
zheng-da commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170129335
##
File path: src/operator/nn/mkldnn/mkldnn_base.cc
##
@@ -270,9 +270,24 @@ void
zheng-da commented on issue #9853: Flaky test_operator.test_binary_op
URL:
https://github.com/apache/incubator-mxnet/issues/9853#issuecomment-367860345
@cjolivier01 in both cases
(https://github.com/apache/incubator-mxnet/issues/9853 and
marcoabreu commented on issue #9853: Flaky test_operator.test_binary_op
URL:
https://github.com/apache/incubator-mxnet/issues/9853#issuecomment-367860051
Well even if it was changed to use MKL, this would not apply here since
we're running on OpenBLAS, right?
nswamy commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367859595
I am ok to put it to all the releases after 0.11.0 upto master, but I don't
think it has to be done right away.
marcoabreu commented on issue #9853: Flaky test_operator.test_binary_op
URL:
https://github.com/apache/incubator-mxnet/issues/9853#issuecomment-367859620
I don't think we have the tools to measure performance on that scale yet. As
far as I know, this is in the works. Since there's still
marcoabreu commented on issue #9843: Flaky
test_random.test_exponential_generator @ Python2: MKLDNN-CPU
URL:
https://github.com/apache/incubator-mxnet/issues/9843#issuecomment-367858507
That might fix the test, but could we risk masking a problem that way? I'm
afraid I'm unfamiliar with
cjolivier01 commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367858332
+1 would have to go into all releases after 0.11
This is an automated message
marcoabreu commented on issue #9856: Flaky test_random.test_gamma_generator
URL:
https://github.com/apache/incubator-mxnet/issues/9856#issuecomment-367857691
No, it's using OpenBLAS.
This is an automated message from the
piiswrong commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367857655
We can't just put stuff into arbitrary previous versions and then have it
disappear later then reappear.
I would just make a fork for this
marcoabreu commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367857486
Contrib also sounds good to me. Are we going to make a formal release or
will that customer just checkout the branch head and compile
nswamy commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367855882
Ok, I registered the symbol under contrib.
This is an automated message from the
marcoabreu commented on issue #9864: Flaky hanging test_operator.test_laop_3 @
Python3: MKLDNN-CPU
URL:
https://github.com/apache/incubator-mxnet/issues/9864#issuecomment-367853511
@asmushetzel would you like to investigate?
marcoabreu opened a new issue #9864: Flaky hanging @ Python3: MKLDNN-CPU
URL: https://github.com/apache/incubator-mxnet/issues/9864
Test times out due to hang.
http://jenkins.mxnet-ci.amazon-ml.com/blue/organizations/jenkins/incubator-mxnet/detail/PR-9848/1/pipeline
```
cjolivier01 commented on issue #9848: add infer_type for regression ops, fix
issues#9847
URL: https://github.com/apache/incubator-mxnet/pull/9848#issuecomment-367853082
Please feel free to add infer type for the other ones if you like (not
required for this PR, however).
marcoabreu commented on issue #9848: add infer_type for regression ops, fix
issues#9847
URL: https://github.com/apache/incubator-mxnet/pull/9848#issuecomment-367852882
Very nice catch, @ZiyueHuang! Sounds like a good idea.
cjolivier01 commented on issue #9848: add infer_type for regression ops, fix
issues#9847
URL: https://github.com/apache/incubator-mxnet/pull/9848#issuecomment-367852782
Makes sense.
This is an automated message from the
marcoabreu commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367850610
But would this re-enabled test have caught this error? Please explain how
we're preventing this problem from
cjolivier01 commented on issue #9853: Flaky test_operator.test_binary_op
URL:
https://github.com/apache/incubator-mxnet/issues/9853#issuecomment-367850176
Wasn't elemwise_add changed to use mkl?
btw, was it verified that mkl is faster for all shapes and types? I saw it
allocates
marcoabreu commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170119050
##
File path: src/ndarray/ndarray.cc
##
@@ -1017,6 +1017,7 @@ inline void
sxjscience commented on issue #9843: Flaky
test_random.test_exponential_generator @ Python2: MKLDNN-CPU
URL:
https://github.com/apache/incubator-mxnet/issues/9843#issuecomment-367849367
As shown in the log, the chi-square value is above the 0.05 threshold
(indicates the check has passed)
sxjscience commented on issue #9843: Flaky
test_random.test_exponential_generator @ Python2: MKLDNN-CPU
URL:
https://github.com/apache/incubator-mxnet/issues/9843#issuecomment-367849367
As shown in the log, the chi-square value is above the 0.05 threshold
(indicates the check has passed)
marcoabreu opened a new pull request #9863: Fix broken website build and update
version number
URL: https://github.com/apache/incubator-mxnet/pull/9863
Update website build script to point to the right file
Add 1.1.0 to build_all_versions.sh
Made build scripts fail fast with set -xe
cjolivier01 commented on a change in pull request #9862: Fix a race condition
in converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170117620
##
File path: tests/python/gpu/test_gluon_model_zoo_gpu.py
##
@@ -37,11
cjolivier01 commented on a change in pull request #9862: Fix a race condition
in converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170117170
##
File path: src/operator/nn/mkldnn/mkldnn_base.cc
##
@@ -270,9 +270,24 @@
cjolivier01 commented on a change in pull request #9862: Fix a race condition
in converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170116831
##
File path: src/operator/nn/mkldnn/mkldnn_base.cc
##
@@ -270,9 +270,24 @@
zheng-da commented on issue #9853: Flaky test_operator.test_binary_op
URL:
https://github.com/apache/incubator-mxnet/issues/9853#issuecomment-367846746
It seems https://github.com/apache/incubator-mxnet/issues/9853 fails in the
same test. It's strange how this fails.
zheng-da commented on issue #9862: Fix a race condition in converting data
layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#issuecomment-367845835
@marcoabreu previously, I disabled the inference test. Now I enabled all
tests. I also added some prints to clearly
zheng-da commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170114168
##
File path: tests/python/gpu/test_gluon_model_zoo_gpu.py
##
@@ -37,11 +37,11
zheng-da commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170114324
##
File path: src/ndarray/ndarray.cc
##
@@ -1017,6 +1017,7 @@ inline void
This is an automated email from the ASF dual-hosted git repository.
marcoabreu pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/master by this push:
new 18816a5 Add CuDNN (v6) support to
marcoabreu closed pull request #9861: Add CuDNN (v6) support to Jetson builds
URL: https://github.com/apache/incubator-mxnet/pull/9861
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a
marcoabreu commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170113340
##
File path: tests/python/gpu/test_gluon_model_zoo_gpu.py
##
@@ -37,11 +37,11
marcoabreu commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170113089
##
File path: src/ndarray/ndarray.cc
##
@@ -1017,6 +1017,7 @@ inline void
marcoabreu commented on a change in pull request #9862: Fix a race condition in
converting data layouts in MKLDNN.
URL: https://github.com/apache/incubator-mxnet/pull/9862#discussion_r170113380
##
File path: tests/python/gpu/test_gluon_model_zoo_gpu.py
##
@@ -105,6 +107,7
zheng-da commented on issue #9843: Flaky test_random.test_exponential_generator
@ Python2: MKLDNN-CPU
URL:
https://github.com/apache/incubator-mxnet/issues/9843#issuecomment-367840947
It seems this problem is related to
https://github.com/apache/incubator-mxnet/issues/9856
zheng-da commented on issue #9856: Flaky test_random.test_gamma_generator
URL:
https://github.com/apache/incubator-mxnet/issues/9856#issuecomment-367839764
Is this problem related to MKLDNN?
This is an automated message from
cjolivier01 commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367833441
"whenever it's convenient"? none of this seems terribly "convenient". Not
sure where that framing came from. Customer obsessions is rarely
marcoabreu commented on issue #9809: fix optimizer bug in CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-367830881
@larroy
This is an automated message from the Apache Git Service.
To
zheng-da opened a new pull request #9862: Fix a race condition in converting
data layouts.
URL: https://github.com/apache/incubator-mxnet/pull/9862
## Description ##
There is a race condition in data layout conversion in the MKLDNN
implementation.
Currently, when NDArray::data() is
sifmelcara commented on issue #9809: fix optimizer bug in CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-367824559
Hmm, in fact I think adding back `static int __make_ ## OptimizerType ## _
## Name ## __ =` is a little bit better because local static
sifmelcara commented on issue #9809: fix optimizer bug in CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-367822502
Hello! Just got a ping from @chsin for this PR. I'm the author of the ODR
fix.
After some git blame, I found this bug was accidentally
sifmelcara commented on issue #9809: fix optimizer bug in CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-367822502
Hello! Just got a ping from @chsin for this PR. I'm the author of the ODR
fix.
After some git blame, I found this bug was accidentally
sifmelcara commented on issue #9809: fix optimizer bug in CPP-Package
URL: https://github.com/apache/incubator-mxnet/pull/9809#issuecomment-367822502
Hello! Just got a ping from @chsin for this PR. I'm the author of the ODR
fix.
After some git blame, I found this bug was accidentally
szha commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367820170
The rules are standards we set for ourselves, and we probably don't want
people to think this is a community that's OK with bending the rules whenever
szha commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367820170
The rules are standards we set for ourselves, and we probably don't want
people to think this is a community that's OK to bend the rules whenever it's
cjolivier01 commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367818944
probably this discussion should occur on dev and see what the mentors think
On Thu, Feb 22, 2018 at 12:42 PM, Marco de Abreu
marcoabreu commented on issue #9858: softsign activation function
URL: https://github.com/apache/incubator-mxnet/pull/9858#issuecomment-367815373
Of course, but we can't guarantee correctness due to lacking CI and we'd
have to make a formal release according to Apache in order to be able
1 - 100 of 159 matches
Mail list logo