[GitHub] piiswrong commented on issue #9958: Parallelization for ROIpooling OP

2018-03-06 Thread GitBox
piiswrong commented on issue #9958: Parallelization for ROIpooling OP URL: https://github.com/apache/incubator-mxnet/pull/9958#issuecomment-370899560 @cjolivier01 This is an automated message from the Apache Git Service. To

[GitHub] szha commented on issue #9986: gluon language modeling dataset and text token reader

2018-03-06 Thread GitBox
szha commented on issue #9986: gluon language modeling dataset and text token reader URL: https://github.com/apache/incubator-mxnet/pull/9986#issuecomment-370906543 Rebased. This is an automated message from the Apache Git

[GitHub] lightingghost commented on issue #9979: Computing the second derivative

2018-03-06 Thread GitBox
lightingghost commented on issue #9979: Computing the second derivative URL: https://github.com/apache/incubator-mxnet/issues/9979#issuecomment-370919814 I will suggest using `autograd.grad` for computing second order derivatives. like

[GitHub] szha commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
szha commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370884321 Haha OK. Jokes aside, did Hang sufficiently address your concern?

[GitHub] piiswrong commented on issue #9982: Unary ops logcdf_normal, derivlogcdf_normal for log CDF of standard n?

2018-03-06 Thread GitBox
piiswrong commented on issue #9982: Unary ops logcdf_normal, derivlogcdf_normal for log CDF of standard n? URL: https://github.com/apache/incubator-mxnet/pull/9982#issuecomment-370896548 Is this exactly the same with scipy.stats.norm.logcdf? If so maybe we call it logcdf too?

[GitHub] aidan-plenert-macdonald commented on issue #9979: Computing the second derivative

2018-03-06 Thread GitBox
aidan-plenert-macdonald commented on issue #9979: Computing the second derivative URL: https://github.com/apache/incubator-mxnet/issues/9979#issuecomment-370912952 @sxjscience If you tell me how, I can help add support for higher order gradients in more operators. A working example and

[GitHub] szha opened a new pull request #10008: gluon dataset download refactor

2018-03-06 Thread GitBox
szha opened a new pull request #10008: gluon dataset download refactor URL: https://github.com/apache/incubator-mxnet/pull/10008 ## Description ## refactor of download logic in #9986 ## Checklist ## ### Essentials ### - [x] Passed code style checking (`make lint`) - [x]

[GitHub] aaronmarkham opened a new pull request #10010: Documentation for build_version_doc scripts folder

2018-03-06 Thread GitBox
aaronmarkham opened a new pull request #10010: Documentation for build_version_doc scripts folder URL: https://github.com/apache/incubator-mxnet/pull/10010 ## Description ## Adding documentation for the `build_version_doc` scripts folder.

[GitHub] szha commented on issue #9986: gluon language modeling dataset and text token reader

2018-03-06 Thread GitBox
szha commented on issue #9986: gluon language modeling dataset and text token reader URL: https://github.com/apache/incubator-mxnet/pull/9986#issuecomment-370887670 I created a separate PR (linked above) for making the download refactoring changes on master, and will rebase this PR once

[incubator-mxnet] branch master updated: [fix issue#9976] The assignment problem in NDArray (#9981)

2018-03-06 Thread jxie
This is an automated email from the ASF dual-hosted git repository. jxie 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 e2b1a56 [fix issue#9976] The assignment

[GitHub] cjolivier01 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
cjolivier01 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370881129 What is the performance impact of these changes for default axes behavior compared ot the older code?

[GitHub] cjolivier01 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
cjolivier01 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370880967 If you have a request for changes from a committer, you can't merge per Apache guidelines.

[GitHub] szha commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
szha commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370881902 Did the guideline explain how committers deal with stale reviews?

[GitHub] yzhliu closed pull request #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
yzhliu closed pull request #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931 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

[incubator-mxnet] branch master updated: Add axes support to Dropout for variational dropout in NLP (#9931)

2018-03-06 Thread liuyizhi
This is an automated email from the ASF dual-hosted git repository. liuyizhi 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 40de6ab Add axes support to Dropout

[GitHub] zhanghang1989 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
zhanghang1989 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370880704 @szha The performance of dropout should be the same as before, when no axes are given.

[GitHub] szha commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
szha commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370881524 Hang suggested that if `axes` is empty, the behavior is exactly the same as before the change

[GitHub] cjolivier01 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
cjolivier01 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370883978 Apache say: "A code-modification proposal may be stopped dead in its tracks by a -1 vote by a qualified

[GitHub] anirudhacharya commented on issue #9963: Onnx Module to import onnx models into mxnet

2018-03-06 Thread GitBox
anirudhacharya commented on issue #9963: Onnx Module to import onnx models into mxnet URL: https://github.com/apache/incubator-mxnet/pull/9963#issuecomment-370888340 @piiswrong can you please take a look at the code. We will keep adding more operator support, but can you review the

[GitHub] piiswrong commented on issue #9991: Random shuffle implementation

2018-03-06 Thread GitBox
piiswrong commented on issue #9991: Random shuffle implementation URL: https://github.com/apache/incubator-mxnet/pull/9991#issuecomment-370892094 Our principle is to try to conform to numpy API as much as possible. Especially in cases like this where users could silently get confusing

[GitHub] piiswrong closed pull request #10008: gluon dataset download refactor

2018-03-06 Thread GitBox
piiswrong closed pull request #10008: gluon dataset download refactor URL: https://github.com/apache/incubator-mxnet/pull/10008 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

[incubator-mxnet] branch master updated: gluon dataset download refactor (#10008)

2018-03-06 Thread jxie
This is an automated email from the ASF dual-hosted git repository. jxie 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 29f7f43 gluon dataset download refactor

[GitHub] larroy opened a new issue #10009: [CMAKE][IoT] Remove pthread from android_arm64 build

2018-03-06 Thread GitBox
larroy opened a new issue #10009: [CMAKE][IoT] Remove pthread from android_arm64 build URL: https://github.com/apache/incubator-mxnet/issues/10009 Logic to detect the target platform should be added to CMake by using try_run and #error as done in openmp

[GitHub] larroy opened a new issue #10011: [CMake][IoT][Jetson] Building with CMake produces libomp.so

2018-03-06 Thread GitBox
larroy opened a new issue #10011: [CMake][IoT][Jetson] Building with CMake produces libomp.so URL: https://github.com/apache/incubator-mxnet/issues/10011 we should use -fopenmp and not 3rd party openmp depending on the target platform

[GitHub] lightingghost opened a new issue #10002: _backward_Convolution is non-differentiable

2018-03-06 Thread GitBox
lightingghost opened a new issue #10002: _backward_Convolution is non-differentiable URL: https://github.com/apache/incubator-mxnet/issues/10002 As I saw mxnet has `autograd` package to support high order gradients, I tried to implement wgan-gp with mxnet. But I got an error of

[GitHub] aidan-plenert-macdonald commented on issue #9979: Computing the second derivative

2018-03-06 Thread GitBox
aidan-plenert-macdonald commented on issue #9979: Computing the second derivative URL: https://github.com/apache/incubator-mxnet/issues/9979#issuecomment-370922133 Gave me ``` MXNetError: [12:44:29] src/pass/gradient.cc:187: Operator _backward_power_scalar is non-differentiable

[GitHub] lightingghost commented on issue #9979: Computing the second derivative

2018-03-06 Thread GitBox
lightingghost commented on issue #9979: Computing the second derivative URL: https://github.com/apache/incubator-mxnet/issues/9979#issuecomment-370919814 I will suggest using `autograd.grad` for computing second order derivatives. like

[GitHub] cjolivier01 closed pull request #9955: [DO NOT MERGE] Cython CI check

2018-03-06 Thread GitBox
cjolivier01 closed pull request #9955: [DO NOT MERGE] Cython CI check URL: https://github.com/apache/incubator-mxnet/pull/9955 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

[GitHub] cjolivier01 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
cjolivier01 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370885831 Yeah, I'm good. This is an automated message

[GitHub] reminisce commented on issue #10006: How to visualise filters of MXNet model?

2018-03-06 Thread GitBox
reminisce commented on issue #10006: How to visualise filters of MXNet model? URL: https://github.com/apache/incubator-mxnet/issues/10006#issuecomment-370885981 Thanks for the reference. We have been working on enabling visualization of MXNet data in the TensorBoard. You can see the

[GitHub] piiswrong closed issue #9976: The assignment problem about NDArray

2018-03-06 Thread GitBox
piiswrong closed issue #9976: The assignment problem about NDArray URL: https://github.com/apache/incubator-mxnet/issues/9976 This is an automated message from the Apache Git Service. To respond to the message, please log on

[GitHub] piiswrong closed pull request #9981: [fix issue#9976] The assignment problem in NDArray

2018-03-06 Thread GitBox
piiswrong closed pull request #9981: [fix issue#9976] The assignment problem in NDArray URL: https://github.com/apache/incubator-mxnet/pull/9981 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

[GitHub] lightingghost commented on issue #10002: _backward_Convolution is non-differentiable

2018-03-06 Thread GitBox
lightingghost commented on issue #10002: _backward_Convolution is non-differentiable URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370905340 @sxjscience I think #9979 is a question rather than a feature request. In his case, he should use `autograd.grad` and

[incubator-mxnet] branch nlp_toolkit updated (02dd89a -> 29f7f43)

2018-03-06 Thread zhasheng
This is an automated email from the ASF dual-hosted git repository. zhasheng pushed a change to branch nlp_toolkit in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 02dd89a Non-blocking row_sparse_pull. Fix incorrect indices generated by device

[GitHub] aidan-plenert-macdonald commented on issue #9979: Computing the second derivative

2018-03-06 Thread GitBox
aidan-plenert-macdonald commented on issue #9979: Computing the second derivative URL: https://github.com/apache/incubator-mxnet/issues/9979#issuecomment-370912952 @sxjscience If you tell me how, I can help add support for higher order gradients in more operators

[GitHub] sxjscience commented on issue #9979: Computing the second derivative

2018-03-06 Thread GitBox
sxjscience commented on issue #9979: Computing the second derivative URL: https://github.com/apache/incubator-mxnet/issues/9979#issuecomment-370914218 @lightingghost Thanks a lot! Could you help give the solution to the question in this PR? I'll double confirm with Eric on the current

[GitHub] aidan-plenert-macdonald commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
aidan-plenert-macdonald commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370929274 I looked at the code and it seems like the problem is that the current implementation is kind of only

[GitHub] mbaijal commented on issue #8913: MXNet 1.0.0 Release Feedback: Source Headers and License Files

2018-03-06 Thread GitBox
mbaijal commented on issue #8913: MXNet 1.0.0 Release Feedback: Source Headers and License Files URL: https://github.com/apache/incubator-mxnet/issues/8913#issuecomment-370929239 All Issues here have been fixed, discussed or tracked in the wiki linked above. I will create new github

[GitHub] mbaijal closed issue #8913: MXNet 1.0.0 Release Feedback: Source Headers and License Files

2018-03-06 Thread GitBox
mbaijal closed issue #8913: MXNet 1.0.0 Release Feedback: Source Headers and License Files URL: https://github.com/apache/incubator-mxnet/issues/8913 This is an automated message from the Apache Git Service. To respond to

[GitHub] szha commented on a change in pull request #9977: Cpu lstm inference

2018-03-06 Thread GitBox
szha commented on a change in pull request #9977: Cpu lstm inference URL: https://github.com/apache/incubator-mxnet/pull/9977#discussion_r172665833 ## File path: python/mxnet/gluon/data/dataloader.py ## @@ -214,6 +214,7 @@ def __iter__(self): worker.start()

[GitHub] sxjscience commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
sxjscience commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370952116 We need to investigate the current OPs and check whether their backward OPs have registered the gradient. Also, we

[GitHub] aaronmarkham commented on issue #10013: update on setting up Scala with MXNet and the IntelliJ IDE

2018-03-06 Thread GitBox
aaronmarkham commented on issue #10013: update on setting up Scala with MXNet and the IntelliJ IDE URL: https://github.com/apache/incubator-mxnet/pull/10013#issuecomment-370982086 Images PR: https://github.com/dmlc/web-data/pull/42

[GitHub] piiswrong commented on issue #10002: _backward_Convolution is non-differentiable

2018-03-06 Thread GitBox
piiswrong commented on issue #10002: _backward_Convolution is non-differentiable URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370925069 2nd order gradient is only implemented for a few operators like * and exp.

[GitHub] sxjscience commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
sxjscience commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370954204 @aidan-plenert-macdonald Yes, we should do that. This

[GitHub] sxjscience commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
sxjscience commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370957305 @aidan-plenert-macdonald I suggest we should first classify the OPs into three categories. - [ ] OPs that we can

[GitHub] aaronmarkham opened a new pull request #10013: update on setting up Scala with MXNet and the IntelliJ IDE

2018-03-06 Thread GitBox
aaronmarkham opened a new pull request #10013: update on setting up Scala with MXNet and the IntelliJ IDE URL: https://github.com/apache/incubator-mxnet/pull/10013 ## Description ## Fixes #9973 Documentation update for MXNet-Scala setup and setting up an IntelliJ project. ##

[GitHub] lightingghost commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
lightingghost commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370984169 @sxjscience I would love to but unfortunately I am not familiar with c++, nor with the mxnet basic architecture. I

[GitHub] sxjscience commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
sxjscience commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370986063 @lightingghost That's okay. You can ask any questions in Github or in (https://discuss.mxnet.io) and we will reply you

[GitHub] mbaijal commented on issue #8920: MXNet 1.0.0 & 1.1.0 Release Feedback: Move the submodules to a separate directory

2018-03-06 Thread GitBox
mbaijal commented on issue #8920: MXNet 1.0.0 & 1.1.0 Release Feedback: Move the submodules to a separate directory URL: https://github.com/apache/incubator-mxnet/issues/8920#issuecomment-370932529 This was requested again in the 1.1.0 Release since the 3000 unlicensed files are mostly

[GitHub] szha commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
szha commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370967526 @zhanghang1989 is working on it This is an automated

[GitHub] lightingghost commented on issue #9979: Computing the second derivative

2018-03-06 Thread GitBox
lightingghost commented on issue #9979: Computing the second derivative URL: https://github.com/apache/incubator-mxnet/issues/9979#issuecomment-370923003 @aidan-plenert-macdonald Yes the similar error I got. It seems second derivative is the feature they are working on.

[GitHub] mbaijal opened a new issue #10012: [Bug] There are two broken links on the top level README.md file which point to the old CI

2018-03-06 Thread GitBox
mbaijal opened a new issue #10012: [Bug] There are two broken links on the top level README.md file which point to the old CI URL: https://github.com/apache/incubator-mxnet/issues/10012 The top level README.md has two links to logos that point to builds.apache.org which is now deprecated.

[GitHub] aidan-plenert-macdonald commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
aidan-plenert-macdonald commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370985296 @sxjscience Sure. I'll take the unary. I'll be working out of

[GitHub] sxjscience commented on issue #10000: fix average pooling kernel size assignment error

2018-03-06 Thread GitBox
sxjscience commented on issue #1: fix average pooling kernel size assignment error URL: https://github.com/apache/incubator-mxnet/pull/1#issuecomment-370974999 @CoinCheung The error is due to that `pool_type` has no default value. We should change this line to default it max

[GitHub] marcoabreu commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
marcoabreu commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370976700 Next time please make sure to have code changes and tests within the same PR instead of splitting them.

[GitHub] sxjscience commented on issue #10000: fix average pooling kernel size assignment error

2018-03-06 Thread GitBox
sxjscience commented on issue #1: fix average pooling kernel size assignment error URL: https://github.com/apache/incubator-mxnet/pull/1#issuecomment-370974999 @CoinCheung The error is due to that `pool_type` has no default value. We should change this line to set the default to

[GitHub] sxjscience commented on issue #10000: fix average pooling kernel size assignment error

2018-03-06 Thread GitBox
sxjscience commented on issue #1: fix average pooling kernel size assignment error URL: https://github.com/apache/incubator-mxnet/pull/1#issuecomment-370974999 @CoinCheung The error is due to that `pool_type` has no default value. We should change this line to set the default to

[GitHub] piiswrong commented on issue #10013: update on setting up Scala with MXNet and the IntelliJ IDE

2018-03-06 Thread GitBox
piiswrong commented on issue #10013: update on setting up Scala with MXNet and the IntelliJ IDE URL: https://github.com/apache/incubator-mxnet/pull/10013#issuecomment-370977991 Please move pics to https://github.com/dmlc/web-data/pulls

[GitHub] sxjscience commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
sxjscience commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370929547 @lightingghost Borrow the discussion in https://github.com/apache/incubator-mxnet/issues/9979 here ```python

[GitHub] sxjscience commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
sxjscience commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370949887 @aidan-plenert-macdonald That's correct, we need to check the backward pass of these operators and register the

[GitHub] marcoabreu commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
marcoabreu commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370963513 Sorry to be a bit late here, but this PR has no test coverage. Could you please elaborate @yzhliu ?

[GitHub] marcoabreu commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
marcoabreu commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370963513 Sorry to be a bit late here, but this PR has no test coverage. Could you please elaborate @yzhliu @szha ?

[GitHub] aidan-plenert-macdonald commented on issue #8667: Compiler/linker error while trying to build from source on Mac OSX Sierra 10.12.6.

2018-03-06 Thread GitBox
aidan-plenert-macdonald commented on issue #8667: Compiler/linker error while trying to build from source on Mac OSX Sierra 10.12.6. URL: https://github.com/apache/incubator-mxnet/issues/8667#issuecomment-370968646 Still a problem on most recent commit as of posting this

[GitHub] aidan-plenert-macdonald commented on issue #8667: Compiler/linker error while trying to build from source on Mac OSX Sierra 10.12.6.

2018-03-06 Thread GitBox
aidan-plenert-macdonald commented on issue #8667: Compiler/linker error while trying to build from source on Mac OSX Sierra 10.12.6. URL: https://github.com/apache/incubator-mxnet/issues/8667#issuecomment-370968646 Still a problem on most recent commit as of posting this ``` commit

[GitHub] Jerryzcn commented on a change in pull request #9977: Cpu lstm inference

2018-03-06 Thread GitBox
Jerryzcn commented on a change in pull request #9977: Cpu lstm inference URL: https://github.com/apache/incubator-mxnet/pull/9977#discussion_r172673076 ## File path: python/mxnet/gluon/data/dataloader.py ## @@ -214,6 +214,7 @@ def __iter__(self):

[GitHub] aidan-plenert-macdonald commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
aidan-plenert-macdonald commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370952929 @sxjscience Would it not be better to register the backward OPs as combinations of the forward ones so

[GitHub] zhanghang1989 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP

2018-03-06 Thread GitBox
zhanghang1989 commented on issue #9931: Add axes support to Dropout for variational dropout in NLP URL: https://github.com/apache/incubator-mxnet/pull/9931#issuecomment-370967897 I am creating another PR for unit test. This

[GitHub] Mabinogiysk commented on issue #10005: Where should I put my dll after building from source?

2018-03-06 Thread GitBox
Mabinogiysk commented on issue #10005: Where should I put my dll after building from source? URL: https://github.com/apache/incubator-mxnet/issues/10005#issuecomment-370986994 I have checked the dependency of `libopenblas.dll` and `opencv_world341.dll`. Only `libgfortran-3.dll` is

[GitHub] sxjscience commented on issue #9979: Computing the second derivative

2018-03-06 Thread GitBox
sxjscience commented on issue #9979: Computing the second derivative URL: https://github.com/apache/incubator-mxnet/issues/9979#issuecomment-370927643 @lightingghost @aidan-plenert-macdonald Let's then redirect it to the other PR.

[GitHub] sxjscience closed issue #9979: Computing the second derivative

2018-03-06 Thread GitBox
sxjscience closed issue #9979: Computing the second derivative URL: https://github.com/apache/incubator-mxnet/issues/9979 This is an automated message from the Apache Git Service. To respond to the message, please log on

[GitHub] sxjscience commented on issue #10002: General support of OPs for second-order gradient

2018-03-06 Thread GitBox
sxjscience commented on issue #10002: General support of OPs for second-order gradient URL: https://github.com/apache/incubator-mxnet/issues/10002#issuecomment-370981003 @lightingghost @aidan-plenert-macdonald Would you have time to work on this? You may refer to the implementation of

[GitHub] anirudh2290 opened a new pull request #10014: Fix crash with mx.nd.ones

2018-03-06 Thread GitBox
anirudh2290 opened a new pull request #10014: Fix crash with mx.nd.ones URL: https://github.com/apache/incubator-mxnet/pull/10014 ## Description ## Reported by @haojin2 Execution of code like the following on device with only 8 gpus will crash the process instead of throwing

[GitHub] GoodJoey opened a new issue #10017: When doing the distributed training, the bandwidth becomes the bottleneck.

2018-03-06 Thread GitBox
GoodJoey opened a new issue #10017: When doing the distributed training, the bandwidth becomes the bottleneck. URL: https://github.com/apache/incubator-mxnet/issues/10017 is there any experience/magic solutions when using mxnet to do distribute training? say, is there any method,

[GitHub] huyangc commented on issue #10017: When doing the distributed training, the bandwidth becomes the bottleneck.

2018-03-06 Thread GitBox
huyangc commented on issue #10017: When doing the distributed training, the bandwidth becomes the bottleneck. URL: https://github.com/apache/incubator-mxnet/issues/10017#issuecomment-371007682 You could try the gradient compression first. It is more easier.

[GitHub] Mabinogiysk commented on issue #10005: Where should I put my dll after building from source?

2018-03-06 Thread GitBox
Mabinogiysk commented on issue #10005: Where should I put my dll after building from source? URL: https://github.com/apache/incubator-mxnet/issues/10005#issuecomment-371007588 I used `MS Process Monitor` tool to monitor the python process while I was importing `mxnet`. I searched the

[GitHub] Mabinogiysk commented on issue #10005: Where should I put my dll after building from source?

2018-03-06 Thread GitBox
Mabinogiysk commented on issue #10005: Where should I put my dll after building from source? URL: https://github.com/apache/incubator-mxnet/issues/10005#issuecomment-371007588 I used `MS Process Monitor` tool to monitor the python process while I was importing `mxnet`. I searched the

[GitHub] CodingCat commented on issue #10018: Bug Fix and performance optimized for rtc

2018-03-06 Thread GitBox
CodingCat commented on issue #10018: Bug Fix and performance optimized for rtc URL: https://github.com/apache/incubator-mxnet/pull/10018#issuecomment-371019866 Hi, the community has passed to vote about associating the code changes with JIRA

[GitHub] CodingCat commented on issue #10010: Documentation for build_version_doc scripts folder

2018-03-06 Thread GitBox
CodingCat commented on issue #10010: Documentation for build_version_doc scripts folder URL: https://github.com/apache/incubator-mxnet/pull/10010#issuecomment-371020112 Hi, the community has passed to vote about associating the code changes with JIRA

[GitHub] CodingCat commented on issue #10014: Fix crash with mx.nd.ones

2018-03-06 Thread GitBox
CodingCat commented on issue #10014: Fix crash with mx.nd.ones URL: https://github.com/apache/incubator-mxnet/pull/10014#issuecomment-37102 Hi, the community has passed to vote about associating the code changes with JIRA

[GitHub] Mabinogiysk commented on issue #9814: C++ Convolution layer usage produces "Segmentation fault"

2018-03-06 Thread GitBox
Mabinogiysk commented on issue #9814: C++ Convolution layer usage produces "Segmentation fault" URL: https://github.com/apache/incubator-mxnet/issues/9814#issuecomment-371033697 Hi, could you tell me where to find the tutorials about building C++ package?

[GitHub] sxjscience commented on a change in pull request #10000: fix average pooling kernel size assignment error

2018-03-06 Thread GitBox
sxjscience commented on a change in pull request #1: fix average pooling kernel size assignment error URL: https://github.com/apache/incubator-mxnet/pull/1#discussion_r172729940 ## File path: tests/python/gpu/test_operator_gpu.py ## @@ -904,86 +904,87 @@ def

[GitHub] CodingCat commented on issue #8107: "add warmup lr_scheduler" create a new pr

2018-03-06 Thread GitBox
CodingCat commented on issue #8107: "add warmup lr_scheduler" create a new pr URL: https://github.com/apache/incubator-mxnet/pull/8107#issuecomment-371021474 Hi, the community has passed to vote about associating the code changes with JIRA

[incubator-mxnet] branch master updated: add SameType as default type inference function in imperative mode (#9942)

2018-03-06 Thread jxie
This is an automated email from the ASF dual-hosted git repository. jxie 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 0fe04e9 add SameType as default type

[GitHub] piiswrong closed pull request #9942: add SameType as default type inference function in imperative mode

2018-03-06 Thread GitBox
piiswrong closed pull request #9942: add SameType as default type inference function in imperative mode URL: https://github.com/apache/incubator-mxnet/pull/9942 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

[GitHub] piiswrong commented on issue #9934: [MXNET-31] Support variable sequence length in gluon.RecurrentCell

2018-03-06 Thread GitBox
piiswrong commented on issue #9934: [MXNET-31] Support variable sequence length in gluon.RecurrentCell URL: https://github.com/apache/incubator-mxnet/pull/9934#issuecomment-371034064 @szha Is this ready? This is an

[GitHub] kandoiNikhil opened a new issue #10016: MxNet hangs up during bind.

2018-03-06 Thread GitBox
kandoiNikhil opened a new issue #10016: MxNet hangs up during bind. URL: https://github.com/apache/incubator-mxnet/issues/10016 ## DataParallelExecutorGroup hangs This is a copy of the resolved issue. See [this](https://github.com/apache/incubator-mxnet/issues/6325) When

[GitHub] CodingCat commented on issue #9877: Better even_split=False support in gluon.split_data()

2018-03-06 Thread GitBox
CodingCat commented on issue #9877: Better even_split=False support in gluon.split_data() URL: https://github.com/apache/incubator-mxnet/pull/9877#issuecomment-371020813 Hi, the community has passed to vote about associating the code changes with JIRA

[GitHub] CodingCat commented on issue #9932: Fixes for profiler

2018-03-06 Thread GitBox
CodingCat commented on issue #9932: Fixes for profiler URL: https://github.com/apache/incubator-mxnet/pull/9932#issuecomment-371020720 Hi, the community has passed to vote about associating the code changes with JIRA

[GitHub] CodingCat commented on issue #9934: [Gluon] Support variable sequence length in gluon.RecurrentCell

2018-03-06 Thread GitBox
CodingCat commented on issue #9934: [Gluon] Support variable sequence length in gluon.RecurrentCell URL: https://github.com/apache/incubator-mxnet/pull/9934#issuecomment-371020687 Hi, the community has passed to vote about associating the code changes with JIRA

[GitHub] CodingCat commented on issue #9933: Adding support to profile kvstore server during distributed training

2018-03-06 Thread GitBox
CodingCat commented on issue #9933: Adding support to profile kvstore server during distributed training URL: https://github.com/apache/incubator-mxnet/pull/9933#issuecomment-371020709 Hi, the community has passed to vote about associating the code changes with JIRA

[GitHub] CodingCat commented on issue #9936: Adjust MKLDNN compile flags order to resolve some corner case

2018-03-06 Thread GitBox
CodingCat commented on issue #9936: Adjust MKLDNN compile flags order to resolve some corner case URL: https://github.com/apache/incubator-mxnet/pull/9936#issuecomment-371020673 Hi, the community has passed to vote about associating the code changes with JIRA

[GitHub] CodingCat commented on issue #9940: remove set CMAKE_GENERATOR_TOOLSET

2018-03-06 Thread GitBox
CodingCat commented on issue #9940: remove set CMAKE_GENERATOR_TOOLSET URL: https://github.com/apache/incubator-mxnet/pull/9940#issuecomment-371020592 Hi, the community has passed to vote about associating the code changes with JIRA

[GitHub] Mabinogiysk opened a new issue #10020: How can I build C++ package?

2018-03-06 Thread GitBox
Mabinogiysk opened a new issue #10020: How can I build C++ package? URL: https://github.com/apache/incubator-mxnet/issues/10020 I can not find anything about installing mxnet for C++. The tutorials about [C++](https://mxnet.incubator.apache.org/tutorials/c++/basics.html) mentions

[GitHub] starimpact commented on issue #8695: Hangs training on P100

2018-03-06 Thread GitBox
starimpact commented on issue #8695: Hangs training on P100 URL: https://github.com/apache/incubator-mxnet/issues/8695#issuecomment-371023082 i encountered the deadlock also mxnet version is v0.8.0 ``` #0 pthread_cond_wait@@GLIBC_2.3.2 () #1 0x7f39f58198dc in

[GitHub] piiswrong commented on issue #10018: Bug Fix and performance optimized for rtc

2018-03-06 Thread GitBox
piiswrong commented on issue #10018: Bug Fix and performance optimized for rtc URL: https://github.com/apache/incubator-mxnet/pull/10018#issuecomment-371026295 @CodingCat When was this? JIRA is ancient and arcane. This only adds unnecessary overhead.

[GitHub] pengzhao-intel commented on issue #9939: add multi proposal operator (cpu version) and fix the bug in proposal op (gpu version)

2018-03-06 Thread GitBox
pengzhao-intel commented on issue #9939: add multi proposal operator (cpu version) and fix the bug in proposal op (gpu version) URL: https://github.com/apache/incubator-mxnet/pull/9939#issuecomment-371026517 @wkcn thanks for the detail profiling data. Regarding sorting, you can use

[GitHub] piiswrong commented on a change in pull request #10014: Fix crash with mx.nd.ones

2018-03-06 Thread GitBox
piiswrong commented on a change in pull request #10014: Fix crash with mx.nd.ones URL: https://github.com/apache/incubator-mxnet/pull/10014#discussion_r172745323 ## File path: src/imperative/imperative_utils.h ## @@ -70,12 +70,14 @@ inline Context GetContext(const

[GitHub] kandoiNikhil commented on issue #6325: DataParallelExecutorGroup hangs

2018-03-06 Thread GitBox
kandoiNikhil commented on issue #6325: DataParallelExecutorGroup hangs URL: https://github.com/apache/incubator-mxnet/issues/6325#issuecomment-370991527 @yajiedesign, would you mind re-opening this issue. Thanks. This is

[GitHub] huyangc commented on issue #10015: Can mxnet support distribute training using nccl v2.0?

2018-03-06 Thread GitBox
huyangc commented on issue #10015: Can mxnet support distribute training using nccl v2.0? URL: https://github.com/apache/incubator-mxnet/issues/10015#issuecomment-371007304 After reading pull requests of [here](https://github.com/apache/incubator-mxnet/pull/8294), I get that the nccl

[GitHub] CodingCat commented on issue #8423: Re-implement segnet in MXnet

2018-03-06 Thread GitBox
CodingCat commented on issue #8423: Re-implement segnet in MXnet URL: https://github.com/apache/incubator-mxnet/pull/8423#issuecomment-371021462 Hi, the community has passed to vote about associating the code changes with JIRA

  1   2   3   >