[GitHub] zhanghang1989 commented on a change in pull request #9688: [MXNET-108] Adding BilinearResize2D and AdaptiveAvgPool2d operators

zhanghang1989 commented on a change in pull request #9688: [MXNET-108] Adding BilinearResize2D and AdaptiveAvgPool2d operators URL: https://github.com/apache/incubator-mxnet/pull/9688#discussion_r179983326 ## File path: src/operator/contrib/adaptive_avg_pooling.cc ## @@

[GitHub] solin319 commented on issue #10366: fix bug in sgd

solin319 commented on issue #10366: fix bug in sgd URL: https://github.com/apache/incubator-mxnet/pull/10366#issuecomment-379620820 The results above was get in multi-GPU training with kv_store='local'. The same problem was in kv_store='device' too. When we training in multi-machine,

[GitHub] solin319 closed pull request #10366: fix bug in sgd

solin319 closed pull request #10366: fix bug in sgd URL: https://github.com/apache/incubator-mxnet/pull/10366 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

[GitHub] szha commented on issue #9686: APIs that might be a good idea to remove in 2.0

szha commented on issue #9686: APIs that might be a good idea to remove in 2.0 URL: https://github.com/apache/incubator-mxnet/issues/9686#issuecomment-379618905 #9881 This is an automated message from the Apache Git

[GitHub] CodingCat opened a new pull request #10462: [MXNET-62] add test against spark integration

CodingCat opened a new pull request #10462: [MXNET-62] add test against spark integration URL: https://github.com/apache/incubator-mxnet/pull/10462 the test would be passed only we 1. review and merge https://github.com/dmlc/ps-lite/pull/133 2. upload the original

[GitHub] hetong007 commented on a change in pull request #10450: [MXNET-212] [R] Fix Random Samplers from Uniform and Gaussian distribution in R bindings

hetong007 commented on a change in pull request #10450: [MXNET-212] [R] Fix Random Samplers from Uniform and Gaussian distribution in R bindings URL: https://github.com/apache/incubator-mxnet/pull/10450#discussion_r179979144 ## File path: R-package/tests/testthat/test_random.R

[GitHub] eric-haibin-lin commented on issue #10388: [MXNET-265] Update optimizer doc to clarify wd behaviors

eric-haibin-lin commented on issue #10388: [MXNET-265] Update optimizer doc to clarify wd behaviors URL: https://github.com/apache/incubator-mxnet/pull/10388#issuecomment-379615395 I'm still hesitated to change all these behaviors since they're will break existing training scripts..

[GitHub] xinyu-intel commented on issue #10365: [MXNET-261]Update MKLDNN & Add CPP Test

xinyu-intel commented on issue #10365: [MXNET-261]Update MKLDNN & Add CPP Test URL: https://github.com/apache/incubator-mxnet/pull/10365#issuecomment-379613321 @nihui Please help take a look at this pr. The gpu unit test **'depth_wise_conv'** which skipped in #10098 can't pass now. Thank

[GitHub] eric-haibin-lin commented on issue #10366: fix bug in sgd

eric-haibin-lin commented on issue #10366: fix bug in sgd URL: https://github.com/apache/incubator-mxnet/pull/10366#issuecomment-379610436 Are you training with multi-machine and multi-GPU? What type of kvstore are you using? I guess one walk-around is to update the `FResourceRequest `

[GitHub] anirudh2290 commented on issue #10354: Expose the number of GPUs.

anirudh2290 commented on issue #10354: Expose the number of GPUs. URL: https://github.com/apache/incubator-mxnet/pull/10354#issuecomment-379596694 @tdomhan any updates ? This is an automated message from the Apache Git

[GitHub] sergeykolychev closed pull request #10460: MXNET-295 [Perl] Added support for linalg operators.

sergeykolychev closed pull request #10460: MXNET-295 [Perl] Added support for linalg operators. URL: https://github.com/apache/incubator-mxnet/pull/10460 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: MXNET-295 [Perl] Added support for linalg operators. (#10460)

This is an automated email from the ASF dual-hosted git repository. sergeykolychev 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 658bad6 MXNET-295 [Perl]

[GitHub] anirudh2290 commented on issue #10000: [MXNET-80] Fix average pooling kernel size assignment error

anirudh2290 commented on issue #1: [MXNET-80] Fix average pooling kernel size assignment error URL: https://github.com/apache/incubator-mxnet/pull/1#issuecomment-379595604 @CoinCheung can you please resolve the conflicts.

[GitHub] anirudh2290 commented on issue #9688: [MXNET-108] Adding BilinearResize2D and AdaptiveAvgPool2d operators

anirudh2290 commented on issue #9688: [MXNET-108] Adding BilinearResize2D and AdaptiveAvgPool2d operators URL: https://github.com/apache/incubator-mxnet/pull/9688#issuecomment-379595224 @zhanghang1989 any updates on this ?

[GitHub] anirudh2290 commented on issue #8797: fix errors for reinterpret_cast usage

anirudh2290 commented on issue #8797: fix errors for reinterpret_cast usage URL: https://github.com/apache/incubator-mxnet/pull/8797#issuecomment-379595162 @xinghedyc any updates ? This is an automated message from the

[GitHub] anirudh2290 commented on issue #8582: Yolo2 operator

anirudh2290 commented on issue #8582: Yolo2 operator URL: https://github.com/apache/incubator-mxnet/pull/8582#issuecomment-379595063 @zhreshold any updates on this ? This is an automated message from the Apache Git Service.

[GitHub] ShootingSpace opened a new pull request #10461: allow user to define unknown token symbol

ShootingSpace opened a new pull request #10461: allow user to define unknown token symbol URL: https://github.com/apache/incubator-mxnet/pull/10461 ## Description ## Add new feature for issue [#10068](https://github.com/apache/incubator-mxnet/issues/10068). It allows unknown token to

[GitHub] ShootingSpace closed pull request #10385: allow user to define unknown token symbol

ShootingSpace closed pull request #10385: allow user to define unknown token symbol URL: https://github.com/apache/incubator-mxnet/pull/10385 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] chengdazhi opened a new issue #9914: Gluon speed issus when input size varies across batches

chengdazhi opened a new issue #9914: Gluon speed issus when input size varies across batches URL: https://github.com/apache/incubator-mxnet/issues/9914 Hi. I have noticed that the gluon framework has speed issues when the input spatial size varies across batches. It causes an approximate

[GitHub] marcoabreu commented on a change in pull request #10435: [MXNET-289] Allow specifying number of batches to run in an epoch to fix hang in dist sync for last epoch

marcoabreu commented on a change in pull request #10435: [MXNET-289] Allow specifying number of batches to run in an epoch to fix hang in dist sync for last epoch URL: https://github.com/apache/incubator-mxnet/pull/10435#discussion_r179961212 ## File path:

[GitHub] sergeykolychev commented on issue #10460: MXNET-295 [Perl] Added support for linalg operators.

sergeykolychev commented on issue #10460: MXNET-295 [Perl] Added support for linalg operators. URL: https://github.com/apache/incubator-mxnet/pull/10460#issuecomment-379581976 @tlby Rob, please check it out This is an

[GitHub] sergeykolychev opened a new pull request #10460: MXNET-295 [Perl] Added support for linalg operators.

sergeykolychev opened a new pull request #10460: MXNET-295 [Perl] Added support for linalg operators. URL: https://github.com/apache/incubator-mxnet/pull/10460 ## Description ## Added support for linear algebra operations on symbols/ndarrays for Perl API ## Checklist ## ###

[GitHub] larroy commented on issue #9808: [MXNET-267] Ssd camera demo

larroy commented on issue #9808: [MXNET-267] Ssd camera demo URL: https://github.com/apache/incubator-mxnet/pull/9808#issuecomment-379579715 @zhreshold thanks for the review. I will look into the video. This is an automated

[GitHub] larroy commented on a change in pull request #9808: [MXNET-267] Ssd camera demo

larroy commented on a change in pull request #9808: [MXNET-267] Ssd camera demo URL: https://github.com/apache/incubator-mxnet/pull/9808#discussion_r179960162 ## File path: example/ssd/demo.py ## @@ -156,3 +209,22 @@ def parse_data_shape(data_shape_str): # run

[GitHub] larroy commented on a change in pull request #9808: [MXNET-267] Ssd camera demo

larroy commented on a change in pull request #9808: [MXNET-267] Ssd camera demo URL: https://github.com/apache/incubator-mxnet/pull/9808#discussion_r179960162 ## File path: example/ssd/demo.py ## @@ -156,3 +209,22 @@ def parse_data_shape(data_shape_str): # run

[GitHub] larroy commented on a change in pull request #9808: [MXNET-267] Ssd camera demo

larroy commented on a change in pull request #9808: [MXNET-267] Ssd camera demo URL: https://github.com/apache/incubator-mxnet/pull/9808#discussion_r179960150 ## File path: example/ssd/demo.py ## @@ -131,13 +138,59 @@ def parse_data_shape(data_shape_str): raise

[GitHub] ShownX commented on issue #9676: [Feature Request] Cross Product Operation

ShownX commented on issue #9676: [Feature Request] Cross Product Operation URL: https://github.com/apache/incubator-mxnet/issues/9676#issuecomment-379578423 Hi, Iacopo @iacopomasi , It is glad to see you are using MXNet now. Cross operation can be implemented by using the current

[GitHub] rahul003 commented on a change in pull request #10435: [MXNET-289] Allow specifying number of batches to run in an epoch to fix hang in dist sync for last epoch

rahul003 commented on a change in pull request #10435: [MXNET-289] Allow specifying number of batches to run in an epoch to fix hang in dist sync for last epoch URL: https://github.com/apache/incubator-mxnet/pull/10435#discussion_r179958709 ## File path:

[GitHub] rahul003 commented on a change in pull request #10435: [MXNET-289] Allow specifying number of batches to run in an epoch to fix hang in dist sync for last epoch

rahul003 commented on a change in pull request #10435: [MXNET-289] Allow specifying number of batches to run in an epoch to fix hang in dist sync for last epoch URL: https://github.com/apache/incubator-mxnet/pull/10435#discussion_r179958665 ## File path:

[GitHub] rahul003 commented on issue #10452: [MXNET-294] Fix element wise multiply for csr ndarrays

rahul003 commented on issue #10452: [MXNET-294] Fix element wise multiply for csr ndarrays URL: https://github.com/apache/incubator-mxnet/pull/10452#issuecomment-379576139 Thanks, added This is an automated message from the

[GitHub] iacopomasi commented on issue #9676: [Feature Request] Cross Product Operation

iacopomasi commented on issue #9676: [Feature Request] Cross Product Operation URL: https://github.com/apache/incubator-mxnet/issues/9676#issuecomment-379575491 Same here I was also looking for cross product operator. Thanks.

[GitHub] anirudh2290 commented on issue #10317: [MXNET-264] Improve performance of MKLDNN in small batch sizes.

anirudh2290 commented on issue #10317: [MXNET-264] Improve performance of MKLDNN in small batch sizes. URL: https://github.com/apache/incubator-mxnet/pull/10317#issuecomment-379573981 @cjolivier01 is this good to merge ?

[GitHub] anirudh2290 commented on issue #10391: [MXNET-139] Tutorial for mixed precision training with float16

anirudh2290 commented on issue #10391: [MXNET-139] Tutorial for mixed precision training with float16 URL: https://github.com/apache/incubator-mxnet/pull/10391#issuecomment-379573420 ping @rahul003 This is an automated

[GitHub] anirudh2290 commented on issue #10183: [MXNET-120] Float16 support for distributed training

anirudh2290 commented on issue #10183: [MXNET-120] Float16 support for distributed training URL: https://github.com/apache/incubator-mxnet/pull/10183#issuecomment-379573117 @rahul003 can you fix the conflicts ? This is

[GitHub] anirudh2290 commented on issue #9152: [MXNET-37] tutorial for distributed training

anirudh2290 commented on issue #9152: [MXNET-37] tutorial for distributed training URL: https://github.com/apache/incubator-mxnet/pull/9152#issuecomment-379572950 @piiswrong can we merge this ? This is an automated message

[GitHub] szha commented on issue #10385: allow user to define unknown token symbol

szha commented on issue #10385: allow user to define unknown token symbol URL: https://github.com/apache/incubator-mxnet/pull/10385#issuecomment-379570094 You can add the test cases to https://github.com/apache/incubator-mxnet/blob/master/tests/python/unittest/test_rnn.py, run it locally

[GitHub] chengdazhi opened a new issue #10459: Does gluon support dynamic graph like if clause?

chengdazhi opened a new issue #10459: Does gluon support dynamic graph like if clause? URL: https://github.com/apache/incubator-mxnet/issues/10459 Hi, I have found that if clause in Block.forward() is executed in main thread and causes delay if outputs of former operators are

[GitHub] chengdazhi closed issue #10240: [Python2] Occasional and random Illegal Memory Access Error | Cached Op Input Context Error

chengdazhi closed issue #10240: [Python2] Occasional and random Illegal Memory Access Error | Cached Op Input Context Error URL: https://github.com/apache/incubator-mxnet/issues/10240 This is an automated message from the

[GitHub] chengdazhi commented on issue #10240: [Python2] Occasional and random Illegal Memory Access Error | Cached Op Input Context Error

chengdazhi commented on issue #10240: [Python2] Occasional and random Illegal Memory Access Error | Cached Op Input Context Error URL: https://github.com/apache/incubator-mxnet/issues/10240#issuecomment-379546948 For future reference, this bug is caused by **CustomOp** for unknown reason.

[GitHub] chengdazhi closed issue #9914: Gluon speed issus when input size varies across batches

chengdazhi closed issue #9914: Gluon speed issus when input size varies across batches URL: https://github.com/apache/incubator-mxnet/issues/9914 This is an automated message from the Apache Git Service. To respond to the

[GitHub] chengdazhi commented on issue #9914: Gluon speed issus when input size varies across batches

chengdazhi commented on issue #9914: Gluon speed issus when input size varies across batches URL: https://github.com/apache/incubator-mxnet/issues/9914#issuecomment-379546718 @marcoabreu Thanks for your prompt reply. Problem is fixed by padding all images to a fixed size. Providing a MVE

[GitHub] marcoabreu commented on issue #10458: [Don't merge] A debug pr of https://github.com/apache/incubator-mxnet/pull/10104

marcoabreu commented on issue #10458: [Don't merge] A debug pr of https://github.com/apache/incubator-mxnet/pull/10104 URL: https://github.com/apache/incubator-mxnet/pull/10458#issuecomment-379538589 In order to reproduce errors from a PR, please use the guide at

[GitHub] marcoabreu closed pull request #10458: [Don't merge] A debug pr of https://github.com/apache/incubator-mxnet/pull/10104

marcoabreu closed pull request #10458: [Don't merge] A debug pr of https://github.com/apache/incubator-mxnet/pull/10104 URL: https://github.com/apache/incubator-mxnet/pull/10458 This is a PR merged from a forked repository. As GitHub hides the original diff on merge, it is displayed

[GitHub] ShootingSpace commented on issue #10385: allow user to define unknown token symbol

ShootingSpace commented on issue #10385: allow user to define unknown token symbol URL: https://github.com/apache/incubator-mxnet/pull/10385#issuecomment-379533858 I actually had a test case, but I have no idea how to call this new feature function since it has not yet merged into mxnet.

[GitHub] lightingghost commented on issue #9676: [Feature Request] Cross Product Operation

lightingghost commented on issue #9676: [Feature Request] Cross Product Operation URL: https://github.com/apache/incubator-mxnet/issues/9676#issuecomment-379530265 How is this going on? I would like to see the cross operator as well =.=

[GitHub] Sherry-Zhang opened a new pull request #10458: A debug pr of pr 10104

Sherry-Zhang opened a new pull request #10458: A debug pr of pr 10104 URL: https://github.com/apache/incubator-mxnet/pull/10458 ## Description ## Just to debug segment fault error in https://github.com/apache/incubator-mxnet/pull/10104, will be closed when bug fixed.

[GitHub] haojin2 commented on issue #10208: [MXNET-117] Sparse operator broadcast_mul/div(csr, dense) = csr

haojin2 commented on issue #10208: [MXNET-117] Sparse operator broadcast_mul/div(csr, dense) = csr URL: https://github.com/apache/incubator-mxnet/pull/10208#issuecomment-379527460 Update on 4/8: A small benchmark for this operator is added to compare the new implementation with the

[GitHub] haojin2 commented on issue #10208: [MXNET-117] Sparse operator broadcast_mul/div(csr, dense) = csr

haojin2 commented on issue #10208: [MXNET-117] Sparse operator broadcast_mul/div(csr, dense) = csr URL: https://github.com/apache/incubator-mxnet/pull/10208#issuecomment-379527460 Update on 4/8: A small benchmark for this operator is added to compare the new implementation with the

[GitHub] jonasrauber commented on issue #6874: mx.symbol.softmax_cross_entropy() raises error "Not enough argument to call operator softmax_cross_entropy"

jonasrauber commented on issue #6874: mx.symbol.softmax_cross_entropy() raises error "Not enough argument to call operator softmax_cross_entropy" URL: https://github.com/apache/incubator-mxnet/issues/6874#issuecomment-379525666 This error is still reproducible!

[GitHub] alues commented on issue #10443: bug fix

alues commented on issue #10443: bug fix URL: https://github.com/apache/incubator-mxnet/pull/10443#issuecomment-379523500 Done This is an automated message from the Apache Git Service. To respond to the message, please log on