This is an automated email from the ASF dual-hosted git repository.
lausen pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from cc78b68 Use relative path for extension unit tests (#17626)
add d2be9a6 Improvements to config.cmake
leezu merged pull request #17639: Set Ninja build generator in config.cmake
instead of commmand line
URL: https://github.com/apache/incubator-mxnet/pull/17639
This is an automated message from the Apache Git Service.
To
apeforest edited a comment on issue #16735: Use single-bit for mask in dropout
operator
URL: https://github.com/apache/incubator-mxnet/pull/16735#issuecomment-589927989
Hi @TaoLv and @PatricZhao I reverted my last commit of "[Do not use
bit-mask when MKL dropout is
apeforest edited a comment on issue #16735: Use single-bit for mask in dropout
operator
URL: https://github.com/apache/incubator-mxnet/pull/16735#issuecomment-589927989
Hi @TaoLv and @PatricZhao I reverted my last commit
apeforest commented on issue #16735: Use single-bit for mask in dropout operator
URL: https://github.com/apache/incubator-mxnet/pull/16735#issuecomment-589927989
Hi @TaoLv I reverted my last commit "
It makes the code too bristle and also involves very complicate logic to
check memory
apeforest edited a comment on issue #16735: Use single-bit for mask in dropout
operator
URL: https://github.com/apache/incubator-mxnet/pull/16735#issuecomment-589927989
Hi @TaoLv and @PatricZhao I reverted my last commit "
It makes the code too bristle and also involves very complicate
This is an automated email from the ASF dual-hosted git repository.
aaronmarkham pushed a commit to branch asf-site
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git
The following commit(s) were added to refs/heads/asf-site by this push:
new e313402 Bump the
This is an automated email from the ASF dual-hosted git repository.
lausen pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from 35083c7 Fix MBCC & scala publish pipeline (#17643)
add cc78b68 Use relative path for extension unit
This is an automated email from the ASF dual-hosted git repository.
lausen pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from 35083c7 Fix MBCC & scala publish pipeline (#17643)
add cc78b68 Use relative path for extension unit
This is an automated email from the ASF dual-hosted git repository.
lausen pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from 376a46f Enable users to specify BLAS (#17648)
add 35083c7 Fix MBCC & scala publish pipeline (#17643)
This is an automated email from the ASF dual-hosted git repository.
lausen pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from 376a46f Enable users to specify BLAS (#17648)
add 35083c7 Fix MBCC & scala publish pipeline (#17643)
leezu merged pull request #17626: Use relative path for extension unit tests
URL: https://github.com/apache/incubator-mxnet/pull/17626
This is an automated message from the Apache Git Service.
To respond to the message,
leezu merged pull request #17643: Fix CI nightly test
URL: https://github.com/apache/incubator-mxnet/pull/17643
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and
haojin2 commented on a change in pull request #17657: Add numpy_fallback
decorator
URL: https://github.com/apache/incubator-mxnet/pull/17657#discussion_r382876918
##
File path: python/mxnet/ndarray/ndarray.py
##
@@ -5087,7 +5087,7 @@ def _make_dl_managed_tensor(array):
leezu commented on issue #17658: [WIP] Update website, README and NEWS with
1.6.0
URL: https://github.com/apache/incubator-mxnet/pull/17658#issuecomment-589902440
@szha what is the status / plan for versioned website?
Currently we don't have a website for MXNet 1.6
This is an automated email from the ASF dual-hosted git repository.
aaronmarkham pushed a commit to branch asf-site
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git
The following commit(s) were added to refs/heads/asf-site by this push:
new 10874ed Bump the
ptrendx opened a new pull request #17658: [WIP] Update website, README and NEWS
with 1.6.0
URL: https://github.com/apache/incubator-mxnet/pull/17658
## Description ##
Update website, README and NEWS with 1.6.0 links.
This
leezu commented on a change in pull request #17657: Add numpy_fallback decorator
URL: https://github.com/apache/incubator-mxnet/pull/17657#discussion_r382868685
##
File path: python/mxnet/ndarray/ndarray.py
##
@@ -5087,7 +5087,7 @@ def _make_dl_managed_tensor(array):
apeforest commented on issue #17599: [Large Tensor] Fixed Embedding op
URL: https://github.com/apache/incubator-mxnet/pull/17599#issuecomment-589894213
@connorgoggins I retriggered CI tests
This is an automated message from
apeforest commented on issue #17639: Set Ninja build generator in config.cmake
instead of commmand line
URL: https://github.com/apache/incubator-mxnet/pull/17639#issuecomment-589885982
@leezu I retriggered CI tests.
This is
haojin2 opened a new pull request #17657: Add numpy_fallback decorator
URL: https://github.com/apache/incubator-mxnet/pull/17657
## Description ##
This decorator is for debug purpose, a user could use this on top of a
function implemented in DeepNumPy and that function will actually run
apeforest commented on issue #17645: [CI] Test CMake builds instead of Makefile
builds
URL: https://github.com/apache/incubator-mxnet/pull/17645#issuecomment-589883136
Right, that makes sense.
This is an automated message
This is an automated email from the ASF dual-hosted git repository.
apeforest pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from 9dcf71d ONNX export: Slice op - Handle None value for ends (#14942)
add 376a46f Enable users to
This is an automated email from the ASF dual-hosted git repository.
apeforest pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git.
from 9dcf71d ONNX export: Slice op - Handle None value for ends (#14942)
add 376a46f Enable users to
apeforest closed pull request #17655: [don't merge] dummy change, ci check
URL: https://github.com/apache/incubator-mxnet/pull/17655
This is an automated message from the Apache Git Service.
To respond to the message, please
apeforest commented on issue #17655: [don't merge] dummy change, ci check
URL: https://github.com/apache/incubator-mxnet/pull/17655#issuecomment-589876750
CI windows-gpu is passing now.
This is an automated message from the
apeforest edited a comment on issue #17655: [don't merge] dummy change, ci check
URL: https://github.com/apache/incubator-mxnet/pull/17655#issuecomment-589876750
CI windows-gpu is passing now. Thanks for the fix.
This is an
apeforest merged pull request #17648: Enable users to specify BLAS
URL: https://github.com/apache/incubator-mxnet/pull/17648
This is an automated message from the Apache Git Service.
To respond to the message, please log on
leezu commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589859111
@icemelon9 please provide the a reproducer to trigger the error message.
This is an automated
leezu commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589858953
@sl1pkn07 please open a separate issue for your problem. This issue is about
MKL.
This is an
samskalicky commented on issue #17623: Dynamic subgraph compile support
URL: https://github.com/apache/incubator-mxnet/pull/17623#issuecomment-589858094
> It sounds like the goal for passing data is to allow data to be compiled
into the bin (example tensort bin) for that subgraph to avoid
sl1pkn07 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589855137
i'm using system openmp and no mkl-dnnl. sorry @icemelon9?
This is an automated message from
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589853130
can you supply a script to reproduc this error? I am not able to reproduce.
This
sl1pkn07 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589852431
and with `readelf -a | grep NEEDED` ?
This is an automated message from the Apache Git
leezu commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589851161
Yes, that's why `dlopen`.
This is an automated message from the Apache Git Service.
To respond
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589849228
Clearly it does not:
[chriso@chriso-ripper:~/src/mxnet/build (master)]ldd
/opt/intel/mkl/lib/intel64/libmkl_rt.so
linux-vdso.so.1
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589848880
Linking in any version of omp statically would probably be a bad idea, since
startup order would be important.
ArmageddonKnight opened a new pull request #17656: [MXNET-1404] Implement
storage tagging, the first half of the memory profiler
URL: https://github.com/apache/incubator-mxnet/pull/17656
## Description ##
implement storage tagging, the first half of the memory profiler
##
leezu edited a comment on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589844040
I think `libmkl_rt` may `dlopen` `libiomp` as per
https://github.com/intel/mkl-dnn/issues/230#issuecomment-451082066, but I
haven't looked into this
leezu commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589844040
I think `libmkl_rt` may `dlopen` `libiomp` as per
https://github.com/intel/mkl-dnn/issues/230#issuecomment-451082066
leezu commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589843442
`libmkl_rt` static links `libiomp` as per
https://github.com/intel/mkl-dnn/issues/230#issuecomment-451073275
leezu removed a comment on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589843442
`libmkl_rt` static links `libiomp` as per
https://github.com/intel/mkl-dnn/issues/230#issuecomment-451073275
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589836549
Actually, i don;t see this behavior when it oes pull in mkl/pulling in the
other omp:
```
[chriso@chriso-ripper:~/src/mxnet/build (master)]ldd
cjolivier01 edited a comment on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589836549
Actually, i don;t see this behavior when it does pull in mkl/pulling in the
other omp (this is Ubuntu 18.04):
```
HahTK commented on issue #17623: Dynamic subgraph compile support
URL: https://github.com/apache/incubator-mxnet/pull/17623#issuecomment-589834860
It sounds like the goal for passing data is to allow data to be compiled
into the bin (example tensort bin) for that subgraph to avoid an init
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589832383
I installed mkl, but it does not appear to pick it up. is there a way to
force it?
This
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589832039
openmp is like a 4-5-second build.
On my desktop machine it's < 3:
real0m2.940s
user0m42.446s
sys 0m5.442s
apeforest commented on issue #17331: [mxnet 2.0] [item 2.4] Turning on large
tensor support by default
URL:
https://github.com/apache/incubator-mxnet/issues/17331#issuecomment-589829033
Thanks to @JonTanS for running the profiler, we have ping pointed the
performance degradation in
ArmageddonKnight commented on issue #14973: [MXNET-1404] Added the GPU memory
profiler
URL: https://github.com/apache/incubator-mxnet/pull/14973#issuecomment-589827768
Thanks to all for the valuable feedbacks. I opened [**another
PR**](https://github.com/apache/incubator-mxnet/pull/17376)
ArmageddonKnight closed pull request #14973: [MXNET-1404] Added the GPU memory
profiler
URL: https://github.com/apache/incubator-mxnet/pull/14973
This is an automated message from the Apache Git Service.
To respond to the
leezu commented on issue #17653: A process crashes when trying to exit normally
URL:
https://github.com/apache/incubator-mxnet/issues/17653#issuecomment-589827963
Could you try deleting the `3rdparty/openmp` directory before configuring
and building. There is an issue where both `libiomp`
eric-haibin-lin commented on issue #17495: Initial inspections for singleton
thread safety in MXNet
URL:
https://github.com/apache/incubator-mxnet/issues/17495#issuecomment-589827805
@leezu thanks for the followup.
This is
leezu commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589827331
> What is pulling in libiomp5.so ?
MKL
> btw, cmake files have min cmake at 3.13, but default 18.04 cmake install
is cmake 3.10. Does anyone know
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589817764
> @cjolivier01 you previously vetoed changing the omp configuration in cmake
build, due to a race condition that had not been fixed. As that has been
cjolivier01 edited a comment on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589812034
btw, cmake files have min cmake at 3.13, but default 18.04 cmake install is
cmake 3.10. Does anyone know what the deal is with 3.13? Ubuntu 18.04
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589812034
btw, cmake files have min cmake at 3.13, but default 18.04 cmake install is
cmake 3.10. Does anyone know what the deal is with 3.13? Ubuntu 18.04 is a
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589809022
1) What is pulling in libiomp5.so ?
2) Since when is libomp being linked in statically? I am not aware of this
ever being the case.
dmitry-markeshov commented on issue #17653: A process crashes when trying to
exit normally
URL:
https://github.com/apache/incubator-mxnet/issues/17653#issuecomment-589808660
> Did you build with cmake or make?
cmake
> Do you have MKL installed on your system?
ChaiBapchya commented on a change in pull request #17642: [OpPerf] Fixed Python
profiler bug
URL: https://github.com/apache/incubator-mxnet/pull/17642#discussion_r382768475
##
File path: benchmark/opperf/utils/profiler_utils.py
##
@@ -248,7 +248,7 @@ def
ChaiBapchya commented on a change in pull request #17642: [OpPerf] Fixed Python
profiler bug
URL: https://github.com/apache/incubator-mxnet/pull/17642#discussion_r382762705
##
File path: benchmark/opperf/utils/profiler_utils.py
##
@@ -248,12 +248,11 @@ def
leezu edited a comment on issue #17645: [CI] Test CMake builds instead of
Makefile builds
URL: https://github.com/apache/incubator-mxnet/pull/17645#issuecomment-589798230
What do you mean? The CI is testing this PR. dev environment is for changes
to the CI itself. This doesn't touch the
ChaiBapchya commented on a change in pull request #17642: [OpPerf] Fixed Python
profiler bug
URL: https://github.com/apache/incubator-mxnet/pull/17642#discussion_r382762557
##
File path: benchmark/opperf/utils/profiler_utils.py
##
@@ -248,12 +248,11 @@ def
leezu commented on issue #17645: [CI] Test CMake builds instead of Makefile
builds
URL: https://github.com/apache/incubator-mxnet/pull/17645#issuecomment-589798230
What do you mean? The CI is testing this PR. This has noting to do with the
dev environment?
apeforest commented on issue #17645: [CI] Test CMake builds instead of Makefile
builds
URL: https://github.com/apache/incubator-mxnet/pull/17645#issuecomment-589791922
Could you first test this in the mxnet-ci-dev environment? If there is
anything wrong, we don't want it to break the CI
leezu commented on issue #17468: Add -march=native -mtune=native to
config/config.cmake
URL: https://github.com/apache/incubator-mxnet/pull/17468#issuecomment-589791317
I plan to merge this after the static build instructions are improved.
leezu commented on issue #17626: Use relative path for extension unit tests
URL: https://github.com/apache/incubator-mxnet/pull/17626#issuecomment-589790854
Retriggered CI
This is an automated message from the Apache Git
leezu commented on issue #17653: A process crashes when trying to exit normally
URL:
https://github.com/apache/incubator-mxnet/issues/17653#issuecomment-589790443
Did you build with cmake or make? Do you have MKL installed on your system?
What's the build configuration? In case it's
apeforest commented on issue #17642: [OpPerf] Fixed Python profiler bug
URL: https://github.com/apache/incubator-mxnet/pull/17642#issuecomment-589789706
> preloaded and multi_* ops aren't being tracked for some reason. could you
fix that too? @connorgoggins
Let's fix that in a
apeforest commented on a change in pull request #17639: Set Ninja build
generator in config.cmake instead of commmand line
URL: https://github.com/apache/incubator-mxnet/pull/17639#discussion_r382740830
##
File path: config/darwin.cmake
##
@@ -27,11 +27,8 @@
# Next
apeforest commented on a change in pull request #17639: Set Ninja build
generator in config.cmake instead of commmand line
URL: https://github.com/apache/incubator-mxnet/pull/17639#discussion_r382740742
##
File path: config/linux.cmake
##
@@ -27,11 +27,8 @@
# Next
apeforest commented on a change in pull request #17639: Set Ninja build
generator in config.cmake instead of commmand line
URL: https://github.com/apache/incubator-mxnet/pull/17639#discussion_r382739586
##
File path: docs/static_site/src/pages/get_started/osx_setup.md
##
apeforest commented on a change in pull request #17639: Set Ninja build
generator in config.cmake instead of commmand line
URL: https://github.com/apache/incubator-mxnet/pull/17639#discussion_r382739707
##
File path: docs/static_site/src/pages/get_started/ubuntu_setup.md
apeforest commented on a change in pull request #17639: Set Ninja build
generator in config.cmake instead of commmand line
URL: https://github.com/apache/incubator-mxnet/pull/17639#discussion_r382739271
##
File path: docs/static_site/src/pages/get_started/osx_setup.md
##
larroy opened a new pull request #17655: dummy change, ci check
URL: https://github.com/apache/incubator-mxnet/pull/17655
## Description ##
CI environment redeploy check.
## Checklist ##
### Essentials ###
Please feel free to remove inapplicable items for your PR.
- [ ]
leezu commented on issue #17648: Enable users to specify BLAS
URL: https://github.com/apache/incubator-mxnet/pull/17648#issuecomment-589768014
Restarted CI job
This is an automated message from the Apache Git Service.
To
connorgoggins commented on issue #17644: [Large Tensor] Fix ravel_multi_index op
URL: https://github.com/apache/incubator-mxnet/pull/17644#issuecomment-589755884
@mxnet-label-bot add [pr-awaiting-review]
This is an automated
leezu commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589750326
@sl1pkn07 given the rapid development of intel-dnnl, MXNet expects a fixed
version of intel-dnnl. It's quite unlikely that the system provides that
particular
leezu commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589749070
@cjolivier01 you previously vetoed changing the omp configuration in cmake
build, due to a race condition that had not been fixed. As that has been fixed,
are
szha commented on a change in pull request #17376: [MXNET-1404] Implement
storage tagging, the first half of the memory profiler
URL: https://github.com/apache/incubator-mxnet/pull/17376#discussion_r382702032
##
File path: include/mxnet/resource.h
##
@@ -148,33 +153,38 @@
sxjscience commented on issue #17654: [LayerNorm] Missing the mismatch cues of
in_channels
URL:
https://github.com/apache/incubator-mxnet/issues/17654#issuecomment-589741619
@ZheyuYe The C++ side implementation of the shape inferring logic is here:
MoisesHer commented on a change in pull request #17652: Use multi-tensor sumSQ
in clip_global_norm
URL: https://github.com/apache/incubator-mxnet/pull/17652#discussion_r382671475
##
File path: python/mxnet/gluon/utils.py
##
@@ -145,12 +143,12 @@ def
apeforest commented on a change in pull request #17652: Use multi-tensor sumSQ
in clip_global_norm
URL: https://github.com/apache/incubator-mxnet/pull/17652#discussion_r382666025
##
File path: python/mxnet/gluon/utils.py
##
@@ -145,12 +143,12 @@ def
ZheyuYe opened a new issue #17654: [LayerNorm] Missing the mismatch cues of
in_channels
URL: https://github.com/apache/incubator-mxnet/issues/17654
## Description
It seems that LayerNorm could work through even the setting of `in_channels`
is wrong. As seen in the reproducible code
dmitry-markeshov opened a new issue #17653: A process crashes when trying to
exit normally
URL: https://github.com/apache/incubator-mxnet/issues/17653
## Description
MxNet version 1.6.1 (manually builded from source)
I run many unit tests through `ctest`. `ctest` calls exec() for
MoisesHer opened a new pull request #17652: Use multi-tensor sumSQ in
clip_global_norm
URL: https://github.com/apache/incubator-mxnet/pull/17652
## Description ##
Using multi-tensor sum of squares in gluon: clip_global_norm.
Instead of computing the sum of squares of each input array
hkvision opened a new issue #17651: Distributed training with kvstore crashes
if worker has different number of data batches
URL: https://github.com/apache/incubator-mxnet/issues/17651
## Description
I'm following here https://mxnet.apache.org/api/faq/distributed_training to
run
sl1pkn07 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589578659
this case also same effects with intel-dnnl.
some distros already provide a package intel-dnnl, but mxnet force download
the sources again
Yiyan66 opened a new pull request #17650: [numpy] add magic methods for bitwise
ops
URL: https://github.com/apache/incubator-mxnet/pull/17650
## Description ##
add magic method
## Checklist ##
### Essentials ###
Please feel free to remove inapplicable items for your PR.
TaoLv commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589568416
Same issue: https://github.com/apache/incubator-mxnet/issues/17366
This is an automated message
sl1pkn07 edited a comment on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589565378
that is because `git clone --recursive` pull openmp, and the cmake script
don't take care if use system openMP or not
sl1pkn07 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-589565378
that is because `git clone --recursive` pull openmp, and the cmake script
don't take care if use system openMP or not
larroy opened a new pull request #17649: add AMI id to instance info on builds
URL: https://github.com/apache/incubator-mxnet/pull/17649
## Description ##
as title
This is an automated message from the Apache Git
leezu commented on a change in pull request #17639: Set Ninja build generator
in config.cmake instead of commmand line
URL: https://github.com/apache/incubator-mxnet/pull/17639#discussion_r382434226
##
File path: docs/static_site/src/pages/get_started/osx_setup.md
##
@@
93 matches
Mail list logo