stu1130 commented on issue #17587: Need help: how to manually build whl
URL:
https://github.com/apache/incubator-mxnet/issues/17587#issuecomment-590510489
[@mxnet-label-bot add question]
This is an automated message from the
stu1130 commented on issue #17664: Question : Bounded delay
URL:
https://github.com/apache/incubator-mxnet/issues/17664#issuecomment-590509979
[@mxnet-label-bot add question]
This is an automated message from the Apache Git
reminisce opened a new pull request #17674: Fix boolean dtype error in
hybridization
URL: https://github.com/apache/incubator-mxnet/pull/17674
## Description ##
Fix https://github.com/apache/incubator-mxnet/issues/17638.
icemelon9 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590506517
I still got the error after tuning off opencv.
```
mxnet git:(master) ✗ readelf -a build/libmxnet.so| grep NEEDED
0x0001 (NEEDED)
mikeobr commented on issue #17139: gamma and beta dimensionality error in
GroupNorm operator
URL:
https://github.com/apache/incubator-mxnet/issues/17139#issuecomment-590488251
Group Norm is available in the latest 1.6 release, does this issue mean it
shouldn't be used?
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 629cac0 Bump the
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590486220
does it occur with opencv turned off?
This is an automated message from the Apache Git
leezu commented on issue #17671: mxnet=1.6.0 for OSX on pypi does not work
URL:
https://github.com/apache/incubator-mxnet/issues/17671#issuecomment-590478397
Seems the linux wheels are labeled wrongly and claim support for all
platforms: https://pypi.org/project/mxnet/1.6.0/#files
leezu removed a comment on issue #17671: mxnet=1.6.0 for OSX on pypi does not
work
URL:
https://github.com/apache/incubator-mxnet/issues/17671#issuecomment-590477816
CC @szha
This is an automated message from the Apache
leezu commented on issue #17671: mxnet=1.6.0 for OSX on pypi does not work
URL:
https://github.com/apache/incubator-mxnet/issues/17671#issuecomment-590477816
CC @szha
This is an automated message from the Apache Git
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590476688
This behavior seems suspicious because the claim would also suggest that the
problem would exist for all similar build cases (ie my machine now) as well
leezu commented on issue #17530: Add deferred compute support
URL: https://github.com/apache/incubator-mxnet/pull/17530#issuecomment-590473083
@eric-haibin-lin enabling deferred compute mode enables recording and
deferring the computation.
It does not automatically export the
icemelon9 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590473106
Yes, I installed Intel MKL 2020.0-166
This is an automated message from the Apache Git
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590472526
This is with latest 2020 version of mkl?
This is an automated message from the Apache Git
icemelon9 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590467401
Here is the cmake log.
```
build git:(master) ✗ cmake -DUSE_CUDA=0 -DUSE_CUDNN=0 -DUSE_MKLDNN=1
-DCMAKE_BUILD_TYPE=Release -GNinja ..
-- The C
leezu edited a comment on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590465449
CI can also reproduce this issue. I switched CI to testing CMake builds
instead of Makefile build in
leezu commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590465449
CI can also reproduce this issue. I switched CI to testing CMake builds
instead of Makefile build in
https://github.com/apache/incubator-mxnet/pull/17645 and
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590450010
```python
(pytorch) [chriso@chriso-ripper:~/src/mxnet (master)]python
Python 3.6.10 |Anaconda, Inc.| (default, Jan 7 2020, 21:14:29)
[GCC 7.3.0]
cjolivier01 commented on issue #17641: OpenMP Error
URL:
https://github.com/apache/incubator-mxnet/issues/17641#issuecomment-590449594
```
0x0001 (NEEDED) Shared library: [libmkl_rt.so]
0x0001 (NEEDED) Shared library: [librt.so.1]
eric-haibin-lin commented on issue #17612: Gluon block thread safety bug
URL:
https://github.com/apache/incubator-mxnet/issues/17612#issuecomment-590430378
This is because self.namescope is shared and overriden:
samskalicky commented on issue #17623: Dynamic subgraph compile support
URL: https://github.com/apache/incubator-mxnet/pull/17623#issuecomment-590388983
@HahTK ive added the ability to pass params down for aux args, and your
suggestion to enable removing the params to the enhancements
kpuatamazon commented on issue #17559: [MXNET-1446] Quantization: intgemm
matrix multiply wrappers
URL: https://github.com/apache/incubator-mxnet/pull/17559#issuecomment-590363395
While I agree with the principle that all operators should be parallel (and
intend to parallelize mine),
rongzha1 commented on issue #17669: flaky test_bf16_operator.test_bf16_bn
URL:
https://github.com/apache/incubator-mxnet/issues/17669#issuecomment-590345377
increase the tolerance for this case in #17673
This is an automated
rongzha1 opened a new pull request #17673: change error tolerance for bf16 bn
test
URL: https://github.com/apache/incubator-mxnet/pull/17673
## Description ##
to fix [issue ](https://github.com/apache/incubator-mxnet/issues/17669),
we enlarge error tolerance to pass bf16 bn test
Alicia1529 opened a new pull request #17672: [Numpy/Don't merge] more support
for boolean indexing/assign and change on implementation
URL: https://github.com/apache/incubator-mxnet/pull/17672
## Description ##
1. implement boolean indexing through advanced indexing by following the
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 8c6c4a3 Bump the
kpuatamazon edited a comment on issue #17559: [MXNET-1446] Quantization:
intgemm matrix multiply wrappers
URL: https://github.com/apache/incubator-mxnet/pull/17559#issuecomment-590256391
@pengzhao-intel Which OMP do you recommend? I've been getting bad OMP
results with a stock install
kpuatamazon commented on issue #17559: [MXNET-1446] Quantization: intgemm
matrix multiply wrappers
URL: https://github.com/apache/incubator-mxnet/pull/17559#issuecomment-590265641
@ciyongch Here is a usage example.
```python
import mxnet as mx
#This is done offline.
weight =
kpuatamazon commented on issue #17559: [MXNET-1446] Quantization: intgemm
matrix multiply wrappers
URL: https://github.com/apache/incubator-mxnet/pull/17559#issuecomment-590256391
@pengzhao-intel Which OMP do you recommend? I've been getting bad OMP
results with a stock install of
pengzhao-intel commented on issue #17669: flaky test_bf16_operator.test_bf16_bn
URL:
https://github.com/apache/incubator-mxnet/issues/17669#issuecomment-590242561
Could we increase the tolerance for this case to align BF16 precision?
fhieber opened a new issue #17671: mxnet=1.6.0 for OSX on pypi does not work
URL: https://github.com/apache/incubator-mxnet/issues/17671
## Description
It seems that mxnet 1.6.0 was published on pypi:
https://pypi.org/project/mxnet/1.6.0/
However, when installing it through pip on a
ElaineBao commented on issue #17669: flaky test_bf16_operator.test_bf16_bn
URL:
https://github.com/apache/incubator-mxnet/issues/17669#issuecomment-590219478
Yes, this happens when the variance of bn meets some specific value which
can not be computed as fp32 by bf16, as you can see,
rongzha1 edited a comment on issue #17669: flaky test_bf16_operator.test_bf16_bn
URL:
https://github.com/apache/incubator-mxnet/issues/17669#issuecomment-590208995
@ElaineBao @ZhennanQin Can you double check it pls? As we discussed, we
can't avoid such cast error. What about skip this
rongzha1 commented on issue #17669: flaky test_bf16_operator.test_bf16_bn
URL:
https://github.com/apache/incubator-mxnet/issues/17669#issuecomment-590208995
@ElaineBao @ZhennanQin Can you double check it pls? As we discussed, we
can't avoid such cast error. What about skip this case ?
ChaiBapchya closed issue #17670: Couldn't load cache Cmake, Ubuntu, Build from
source
URL: https://github.com/apache/incubator-mxnet/issues/17670
This is an automated message from the Apache Git Service.
To respond to the
ChaiBapchya commented on issue #17670: Couldn't load cache Cmake, Ubuntu, Build
from source
URL:
https://github.com/apache/incubator-mxnet/issues/17670#issuecomment-590204046
Got it resolved by removing the CMakeLists.txt file
Likely caused by some corrupt cmake file
ChaiBapchya opened a new issue #17670: Couldn't load cache Cmake, Ubuntu, Build
from source
URL: https://github.com/apache/incubator-mxnet/issues/17670
## Description
Ubuntu16.04 Build from source using Cmake issue
### Error Message
```
Error: could not load cache
```
ChaiBapchya commented on issue #17670: Couldn't load cache Cmake, Ubuntu, Build
from source
URL:
https://github.com/apache/incubator-mxnet/issues/17670#issuecomment-590203716
@mxnet-label-bot add [build, cmake]
This is an
101 - 138 of 138 matches
Mail list logo
