nswamy opened a new pull request #14759: Fix softmax behavior to not cast up
the accumulator if no output dtype is specified
URL: https://github.com/apache/incubator-mxnet/pull/14759
## Description ##
Softmax operator was recently changed to cast the accumulator type to higher
precision
reminisce opened a new pull request #14758: [WIP] Infra for supporting numpy
ops in imperative mode and Gluon APIs
URL: https://github.com/apache/incubator-mxnet/pull/14758
## Description ##
(Brief description on what this PR is about)
## Checklist ##
### Essentials ###
Plea
larroy opened a new pull request #14757: Improve CMake handling of sse2 and sse3
URL: https://github.com/apache/incubator-mxnet/pull/14757
Fix #14304 where CFLAGS were overriden instead of appended to.
## Description ##
This fixes a problem when compiling when F16C is enabled but t
mxnet-label-bot commented on issue #14756: Mismatch between shape Java API
URL:
https://github.com/apache/incubator-mxnet/issues/14756#issuecomment-485325693
Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest some labels so
that the appropriat
androuino opened a new issue #14756: Mismatch between shape Java API
URL: https://github.com/apache/incubator-mxnet/issues/14756
Hi, I’ve finished training the yolo3_darknet53 params and wanted to test it
with Java API, however am having this error which I have no idea where it
coming from
mxnet-label-bot commented on issue #14755: Build on Linux will fail if "test"
option used
URL:
https://github.com/apache/incubator-mxnet/issues/14755#issuecomment-485314106
Hey, this is the MXNet Label Bot.
Thank you for submitting the issue! I will try and suggest some labels so
tha
juliusshufan opened a new issue #14755: Build on Linux will fail if "test"
option used
URL: https://github.com/apache/incubator-mxnet/issues/14755
## Description
There is a build issue happened on MXNet github repo on Linux if using below
command:
make -j USE_OPENCV=1 USE_MKLDNN=1 U
stereomatchingkiss commented on issue #14619: [Discussion] 1.5.0 Roadmap
URL:
https://github.com/apache/incubator-mxnet/issues/14619#issuecomment-485300652
Any plan to simplify compilation process on windows?
Any document to show us how to compile mxnet with the support of mkldnn on
win
This is an automated email from the ASF dual-hosted git repository.
zhasheng 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 2e898ea Bump the publish ti
haojin2 opened a new pull request #14754: [DO NOT REVIEW] Numpy unary functions
URL: https://github.com/apache/incubator-mxnet/pull/14754
## Description ##
As title.
## Checklist ##
### Essentials ###
Please feel free to remove inapplicable items for your PR.
- [ ] The PR
szha commented on a change in pull request #14208: Add support for fast
variable-length LSTM
URL: https://github.com/apache/incubator-mxnet/pull/14208#discussion_r277183673
##
File path: python/mxnet/gluon/rnn/rnn_layer.py
##
@@ -214,7 +215,7 @@ def begin_state(self, batch
Chouffe commented on issue #14752: [Clojure] Better api docstrings by replacing
newlines
URL: https://github.com/apache/incubator-mxnet/pull/14752#issuecomment-485286706
LGTM! Not sure why the windows-cpu CI job is failing though. :100:
@kedarbellare!
-
Chouffe commented on a change in pull request #14750: [clojure][generator][WIP]
add random namespace generation
URL: https://github.com/apache/incubator-mxnet/pull/14750#discussion_r277183497
##
File path: contrib/clojure-package/src/dev/generator.clj
##
@@ -470,6 +476,53
Chouffe commented on a change in pull request #14750: [clojure][generator][WIP]
add random namespace generation
URL: https://github.com/apache/incubator-mxnet/pull/14750#discussion_r277183483
##
File path: contrib/clojure-package/src/dev/generator.clj
##
@@ -470,6 +476,53
KellenSunderland edited a comment on issue #14619: [Discussion] 1.5.0 Roadmap
URL:
https://github.com/apache/incubator-mxnet/issues/14619#issuecomment-485279772
Features I'd like to see for 1.5 include:
Amp if ready
- https://github.com/apache/incubator-mxnet/pull/14173
-
KellenSunderland removed a comment on issue #14619: [Discussion] 1.5.0 Roadmap
URL:
https://github.com/apache/incubator-mxnet/issues/14619#issuecomment-485281604
For those doing the license review I also had a quick question. Would it be
easier if the TensorRT source dependency was made a
KellenSunderland commented on issue #14619: [Discussion] 1.5.0 Roadmap
URL:
https://github.com/apache/incubator-mxnet/issues/14619#issuecomment-485281604
For those doing the license review I also had a quick question. Would it be
easier if the TensorRT source dependency was made a build-c
KellenSunderland edited a comment on issue #14619: [Discussion] 1.5.0 Roadmap
URL:
https://github.com/apache/incubator-mxnet/issues/14619#issuecomment-485279772
Features I'd like to see for 1.5 include:
Amp if ready
- https://github.com/apache/incubator-mxnet/pull/14173
-
This is an automated email from the ASF dual-hosted git repository.
zhasheng 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 955f1e0 Bump the publish ti
KellenSunderland edited a comment on issue #14619: [Discussion] 1.5.0 Roadmap
URL:
https://github.com/apache/incubator-mxnet/issues/14619#issuecomment-485279772
Features I'd like to see for 1.5 include:
Amp if ready
- https://github.com/apache/incubator-mxnet/pull/14173
-
KellenSunderland commented on issue #14159: [Feature Request] Support fp16 for
c++ api
URL:
https://github.com/apache/incubator-mxnet/issues/14159#issuecomment-485281091
I can give a quick summary of my experience with fp16 and C++ so far:
I believe the best way to do this is as in
This is an automated email from the ASF dual-hosted git repository.
reminisce pushed a commit to branch numpy
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/numpy by this push:
new bb3afd4 [Do not review] [Do not merge
reminisce merged pull request #14739: [Do not review] [Do not merge] New
numpy-compatible sum
URL: https://github.com/apache/incubator-mxnet/pull/14739
This is an automated message from the Apache Git Service.
To respond to
KellenSunderland commented on issue #14619: [Discussion] 1.5.0 Roadmap
URL:
https://github.com/apache/incubator-mxnet/issues/14619#issuecomment-485279772
Features I'd like to see for 1.5 include:
* Amp if ready.
* New TensorRT integration with subgraph support
* NVTX ranges for
This is an automated email from the ASF dual-hosted git repository.
zhasheng 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 bec56f9 Bump the publish ti
arcadiaphy opened a new pull request #14753: fix custom op fork test
URL: https://github.com/apache/incubator-mxnet/pull/14753
## Description ##
The custom op fork test introduced in #14451 will cause error when running
with gpu. The common situation is:
```
test_operator_gpu.test_
pengwangucla edited a comment on issue #8458: Synchronized Batch normalization
in multi GPUs training
URL:
https://github.com/apache/incubator-mxnet/issues/8458#issuecomment-485266722
> The SyncBN does reduce the speed due to synchronization for every BN layer
during both forward and
pengwangucla commented on issue #8458: Synchronized Batch normalization in
multi GPUs training
URL:
https://github.com/apache/incubator-mxnet/issues/8458#issuecomment-485266722
I understand it reduces speed, but it also reduces my accuracy, which makes
me really confused.
> The Syn
arcadiaphy commented on issue #14745: fix min max on zero-sized ndarray
URL: https://github.com/apache/incubator-mxnet/pull/14745#issuecomment-485265028
An unrelated custom operator error happens in CI, I'll create another PR to
address it.
-
shoubhik commented on a change in pull request #14749: [Mxnet-1397] Support
symbolic api for requantize and dequantize
URL: https://github.com/apache/incubator-mxnet/pull/14749#discussion_r277173506
##
File path: src/operator/quantization/requantize.cc
##
@@ -61,6 +61,10 @
shoubhik commented on a change in pull request #14749: [Mxnet-1397] Support
symbolic api for requantize and dequantize
URL: https://github.com/apache/incubator-mxnet/pull/14749#discussion_r277173496
##
File path: src/operator/quantization/dequantize.cc
##
@@ -84,6 +84,10 @
SSE4 commented on issue #13400: [MXNET-1229] use OpenBLAS, lapack & OpenCV from
conan
URL: https://github.com/apache/incubator-mxnet/pull/13400#issuecomment-485258988
@Roshrini okay, I'll revive new mail thread with votes tomorrow
---
TaoLv commented on issue #14608: [WIP]Refactor requantize
URL: https://github.com/apache/incubator-mxnet/pull/14608#issuecomment-485252704
@ZhennanQin please take a look at the conflicts and CI failures. Thank you.
This is an
TaoLv commented on a change in pull request #14713: MKLDNN RNN Inference
Integration(fp32 LSTM and vRNN with tanh and relu)
URL: https://github.com/apache/incubator-mxnet/pull/14713#discussion_r277167970
##
File path: src/operator/nn/mkldnn/mkldnn_rnn_impl.h
##
@@ -0,0 +1,
TaoLv commented on a change in pull request #14713: MKLDNN RNN Inference
Integration(fp32 LSTM and vRNN with tanh and relu)
URL: https://github.com/apache/incubator-mxnet/pull/14713#discussion_r277168155
##
File path: src/operator/nn/mkldnn/mkldnn_rnn_impl.h
##
@@ -0,0 +1,
TaoLv commented on a change in pull request #14713: MKLDNN RNN Inference
Integration(fp32 LSTM and vRNN with tanh and relu)
URL: https://github.com/apache/incubator-mxnet/pull/14713#discussion_r277168079
##
File path: src/operator/nn/mkldnn/mkldnn_rnn_impl.h
##
@@ -0,0 +1,
TaoLv commented on a change in pull request #14713: MKLDNN RNN Inference
Integration(fp32 LSTM and vRNN with tanh and relu)
URL: https://github.com/apache/incubator-mxnet/pull/14713#discussion_r277167927
##
File path: src/operator/nn/mkldnn/mkldnn_rnn_impl.h
##
@@ -0,0 +1,
TaoLv commented on a change in pull request #14713: MKLDNN RNN Inference
Integration(fp32 LSTM and vRNN with tanh and relu)
URL: https://github.com/apache/incubator-mxnet/pull/14713#discussion_r277168188
##
File path: src/operator/nn/mkldnn/mkldnn_rnn_impl.h
##
@@ -0,0 +1,
This is an automated email from the ASF dual-hosted git repository.
zhasheng 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 fec89fd Bump the publish ti
TaoLv commented on a change in pull request #14749: [Mxnet-1397] Support
symbolic api for requantize and dequantize
URL: https://github.com/apache/incubator-mxnet/pull/14749#discussion_r277167425
##
File path: src/operator/quantization/dequantize.cc
##
@@ -84,6 +84,10 @@ b
TaoLv commented on a change in pull request #14749: [Mxnet-1397] Support
symbolic api for requantize and dequantize
URL: https://github.com/apache/incubator-mxnet/pull/14749#discussion_r277167427
##
File path: src/operator/quantization/requantize.cc
##
@@ -61,6 +61,10 @@ i
TaoLv commented on issue #14743: change mxnet_option behavior
URL: https://github.com/apache/incubator-mxnet/pull/14743#issuecomment-485249309
#14740 is merged now. Can you rebase? @yajiedesign
This is an automated message fr
This is an automated email from the ASF dual-hosted git repository.
taolv 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 100586a Change size_t to int within for
TaoLv commented on issue #14740: Change size_t to int within for loop to fix
windows build error
URL: https://github.com/apache/incubator-mxnet/pull/14740#issuecomment-485249215
Thank you for the contribution @ciyongch. It's now merged.
-
TaoLv merged pull request #14740: Change size_t to int within for loop to fix
windows build error
URL: https://github.com/apache/incubator-mxnet/pull/14740
This is an automated message from the Apache Git Service.
To respond
ciyongch commented on a change in pull request #14740: Change size_t to int
within for loop to fix windows build error
URL: https://github.com/apache/incubator-mxnet/pull/14740#discussion_r277160322
##
File path: src/operator/subgraph/mkldnn/mkldnn_fc.cc
##
@@ -156,7 +156,
This is an automated email from the ASF dual-hosted git repository.
zhasheng 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 dc1ad1c Bump the publish ti
47 matches
Mail list logo