[GitHub] cjolivier01 commented on a change in pull request #10139: [REQUEST FOR REVIEW] [MXNET-109] Logging APIs for Visualizing MXNet Data in TensorBoard

cjolivier01 commented on a change in pull request #10139: [REQUEST FOR REVIEW] [MXNET-109] Logging APIs for Visualizing MXNet Data in TensorBoard URL: https://github.com/apache/incubator-mxnet/pull/10139#discussion_r175652651 ## File path: tests/python/unittest/test_summary.py

[GitHub] juliusshufan opened a new issue #10168: CPP test case fails to compile using gcc 4.8.5 when MKLDNN enabled

juliusshufan opened a new issue #10168: CPP test case fails to compile using gcc 4.8.5 when MKLDNN enabled URL: https://github.com/apache/incubator-mxnet/issues/10168 ## Description When compiling the CPP test cases using GCC 4.8.5 with MKLDNN enabled, a compiling error will be raised

[GitHub] reminisce commented on issue #10158: Assertion fail when creating very large matrix

reminisce commented on issue #10158: Assertion fail when creating very large matrix URL: https://github.com/apache/incubator-mxnet/issues/10158#issuecomment-374462741 As far as I know, seems no easy workaround right now. May I know in what case you would need to allocate such a big

[GitHub] szha commented on issue #9614: MobileNetV2

szha commented on issue #9614: MobileNetV2 URL: https://github.com/apache/incubator-mxnet/pull/9614#issuecomment-374462384 @Jing-Luo It's currently WIP. This is an automated message from the Apache Git Service. To respond to

[GitHub] Jing-Luo commented on issue #9614: MobileNetV2

Jing-Luo commented on issue #9614: MobileNetV2 URL: https://github.com/apache/incubator-mxnet/pull/9614#issuecomment-374461444 Any plan to release the pretrained model? I try to train the model by my own and it looks not so well... For more details, please look at

[GitHub] reminisce commented on a change in pull request #10139: [REQUEST FOR REVIEW] [MXNET-109] Logging APIs for Visualizing MXNet Data in TensorBoard

reminisce commented on a change in pull request #10139: [REQUEST FOR REVIEW] [MXNET-109] Logging APIs for Visualizing MXNet Data in TensorBoard URL: https://github.com/apache/incubator-mxnet/pull/10139#discussion_r175646730 ## File path: python/setup.py ## @@ -99,6 +99,18

[GitHub] sxjscience opened a new issue #10167: HybridBlock can be slower than Block

sxjscience opened a new issue #10167: HybridBlock can be slower than Block URL: https://github.com/apache/incubator-mxnet/issues/10167 Here's an interesting example that HybridBlock can sometimes be much slower than Block. We define an Identity block using two strategies and compare the

[GitHub] zhiz21 commented on issue #8715: Documentation for MultiBoxPrior is not substantial

zhiz21 commented on issue #8715: Documentation for MultiBoxPrior is not substantial URL: https://github.com/apache/incubator-mxnet/issues/8715#issuecomment-374459578 proposed labels: "Python", "Doc", "Unclear Error/Doc"

[GitHub] zhiz21 commented on issue #9110: 3D dilation support

zhiz21 commented on issue #9110: 3D dilation support URL: https://github.com/apache/incubator-mxnet/issues/9110#issuecomment-374459176 proposed labels: "Bug", "Python", "Gluon", "CUDA" This is an automated message from the

[GitHub] zhiz21 commented on issue #9167: Gluon raises error if the user does not call nd.waitall()

zhiz21 commented on issue #9167: Gluon raises error if the user does not call nd.waitall() URL: https://github.com/apache/incubator-mxnet/issues/9167#issuecomment-374458386 proposed labels: "Gluon", "Python", "Bug" This is

[GitHub] anirudhacharya commented on issue #10145: Inconsistent outputs with mxnet symbol.softmax vs ndarray.softmax

anirudhacharya commented on issue #10145: Inconsistent outputs with mxnet symbol.softmax vs ndarray.softmax URL: https://github.com/apache/incubator-mxnet/issues/10145#issuecomment-374458182 The mistake was in my code. In the following part, I should not have enclosed val in parentheses.

[GitHub] anirudhacharya closed issue #10145: Inconsistent outputs with mxnet symbol.softmax vs ndarray.softmax

anirudhacharya closed issue #10145: Inconsistent outputs with mxnet symbol.softmax vs ndarray.softmax URL: https://github.com/apache/incubator-mxnet/issues/10145 This is an automated message from the Apache Git Service. To

[GitHub] anirudhacharya closed issue #10154: Symbol API's Concat operator inconsistent with NDArray API's Concat operator

anirudhacharya closed issue #10154: Symbol API's Concat operator inconsistent with NDArray API's Concat operator URL: https://github.com/apache/incubator-mxnet/issues/10154 This is an automated message from the Apache Git

[GitHub] zhiz21 commented on issue #7184: Is there any monitor tool for cluster system resources ?

zhiz21 commented on issue #7184: Is there any monitor tool for cluster system resources ? URL: https://github.com/apache/incubator-mxnet/issues/7184#issuecomment-374457618 proposed labels: "Question", "Discussion", "Feature request"

[GitHub] PaulaQin opened a new issue #10166: Please help to provide mxnet 1.1.0 python package for Windows

PaulaQin opened a new issue #10166: Please help to provide mxnet 1.1.0 python package for Windows URL: https://github.com/apache/incubator-mxnet/issues/10166 Hi, Would you please help to build latest version mxnet python package for Windows on PyPI? I want to run mxnet on my Windows

[GitHub] anirudhacharya commented on issue #10154: Symbol API's Concat operator inconsistent with NDArray API's Concat operator

anirudhacharya commented on issue #10154: Symbol API's Concat operator inconsistent with NDArray API's Concat operator URL: https://github.com/apache/incubator-mxnet/issues/10154#issuecomment-374457561 @eric-haibin-lin Correct, I made a mistake in my code. Thanks for correcting me.

[GitHub] cjolivier01 commented on issue #10161: Fix mem alignment.

cjolivier01 commented on issue #10161: Fix mem alignment. URL: https://github.com/apache/incubator-mxnet/pull/10161#issuecomment-374457241 good catch! ok, so then needs to be a constant declaration somewhere — constexpr This

[GitHub] zhiz21 commented on issue #7895: how can i keep the width and height same after Conv2d layer

zhiz21 commented on issue #7895: how can i keep the width and height same after Conv2d layer URL: https://github.com/apache/incubator-mxnet/issues/7895#issuecomment-374456163 proposed labels: "Question", "HowTo" Can you please ask usage related questions at - https://discuss.mxnet.io/

[GitHub] TaoLv commented on issue #10161: Fix mem alignment.

TaoLv commented on issue #10161: Fix mem alignment. URL: https://github.com/apache/incubator-mxnet/pull/10161#issuecomment-374452345 Also need change here? https://github.com/apache/incubator-mxnet/blob/master/src/operator/nn/mkldnn/mkldnn_base-inl.h#L230

[GitHub] anirudh2290 commented on issue #7533: Call symbol/nd.smooth_l1 without scalar causing uncaught exception

anirudh2290 commented on issue #7533: Call symbol/nd.smooth_l1 without scalar causing uncaught exception URL: https://github.com/apache/incubator-mxnet/issues/7533#issuecomment-374451610 @sxjscience Please also tag as "Bug" because of the crash.

[GitHub] anirudh2290 commented on issue #9475: OpenCV Error: Assertion failed (dst.cols < SHRT_MAX && dst.rows < SHRT_MAX && src.cols < SHRT_MAX && src.rows < SHRT_MAX) in remap, file /home/travis/bui

anirudh2290 commented on issue #9475: OpenCV Error: Assertion failed (dst.cols < SHRT_MAX && dst.rows < SHRT_MAX && src.cols < SHRT_MAX && src.rows < SHRT_MAX) in remap, file /home/travis/build/dmlc/mxnet-distro/deps/opencv-3.3.0/modules/imgproc/src/imgwarp.cpp, line 4944 terminate called

[GitHub] anirudh2290 commented on issue #8663: terminate called after throwing an instance of 'std::out_of_range’

anirudh2290 commented on issue #8663: terminate called after throwing an instance of 'std::out_of_range’ URL: https://github.com/apache/incubator-mxnet/issues/8663#issuecomment-374451287 @eric-haibin-lin also, please tag "Bug" for the crash.

[GitHub] anirudh2290 commented on issue #9913: TODO list for Exception Handling Support

anirudh2290 commented on issue #9913: TODO list for Exception Handling Support URL: https://github.com/apache/incubator-mxnet/issues/9913#issuecomment-374451026 Should fix: #9475, #8663 This is an automated message from the

[GitHub] anirudh2290 commented on issue #8749: [FIX ATTACHED] Not working example on Iterators tutorial - SimpleIter

anirudh2290 commented on issue #8749: [FIX ATTACHED] Not working example on Iterators tutorial - SimpleIter URL: https://github.com/apache/incubator-mxnet/issues/8749#issuecomment-374447522 @pracheer can you please verify and close this if this is fixed ?

[GitHub] ThomasDelteil opened a new pull request #10165: Add the ability to exclude specific lines in tutorial notebooks generated from .md

ThomasDelteil opened a new pull request #10165: Add the ability to exclude specific lines in tutorial notebooks generated from .md URL: https://github.com/apache/incubator-mxnet/pull/10165 ## Description ## In tutorials files, sometimes we have statements like `img.plot()` that

[GitHub] pengzhao-intel commented on issue #10161: Fix mem alignment.

pengzhao-intel commented on issue #10161: Fix mem alignment. URL: https://github.com/apache/incubator-mxnet/pull/10161#issuecomment-374442880 @zheng-da please run the benchmark_score.py to confirm the performance impact of mem changes.

[GitHub] anirudh2290 commented on issue #8801: Cannot find custom operator when visualizing network.

anirudh2290 commented on issue #8801: Cannot find custom operator when visualizing network. URL: https://github.com/apache/incubator-mxnet/issues/8801#issuecomment-374442466 @LeonJWH can you confirm that you could fix this after @PhanTask 's suggestion?

[GitHub] ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374411651 I see how that could be a problem, but I think that philosophically MXNet should let the users choose when to block. Typically I

[GitHub] anirudh2290 commented on issue #8914: The custom operator not supported for group context?

anirudh2290 commented on issue #8914: The custom operator not supported for group context? URL: https://github.com/apache/incubator-mxnet/issues/8914#issuecomment-374440415 @mg0880gm are you still seeing this issue with the latest version of mxnet ?

[GitHub] tianweiz07 opened a new issue #10164: unit tests fail on RaspBerry pi with MxNet 1.0

tianweiz07 opened a new issue #10164: unit tests fail on RaspBerry pi with MxNet 1.0 URL: https://github.com/apache/incubator-mxnet/issues/10164 ## Description Install MxNet 1.0 on RaspBerry pi using docker_multiarch. When running uni tests, some of them fail ## Environment

[GitHub] anirudh2290 commented on issue #9613: Transition matrix is not updated during the train in the lstm_crf example

anirudh2290 commented on issue #9613: Transition matrix is not updated during the train in the lstm_crf example URL: https://github.com/apache/incubator-mxnet/issues/9613#issuecomment-374438126 @zhangchen-qinyinghua i cannot see your reviews in #7253 . can you please provide a minimum

[GitHub] nicklhy commented on issue #10148: How to apply the math operators to NDArray in C++

nicklhy commented on issue #10148: How to apply the math operators to NDArray in C++ URL: https://github.com/apache/incubator-mxnet/issues/10148#issuecomment-374435949 Hope we can make a change to it soon. 囧 This is an

[GitHub] rahul003 commented on issue #9028: How to use the C API to train simple a Softmax in pure C?

rahul003 commented on issue #9028: How to use the C API to train simple a Softmax in pure C? URL: https://github.com/apache/incubator-mxnet/issues/9028#issuecomment-374305166 @cjolivier01 Could you please add labels Question

[GitHub] rahul003 commented on issue #9028: How to use the C API to train simple a Softmax in pure C?

rahul003 commented on issue #9028: How to use the C API to train simple a Softmax in pure C? URL: https://github.com/apache/incubator-mxnet/issues/9028#issuecomment-374305166 @cjolivier01 Could you please add the label Question

[GitHub] Roshrini commented on issue #10163: Set threads created by MXNetHandler to Daemon threads so they exit when the main thread exits.

Roshrini commented on issue #10163: Set threads created by MXNetHandler to Daemon threads so they exit when the main thread exits. URL: https://github.com/apache/incubator-mxnet/pull/10163#issuecomment-374435841 Lgtm This

[GitHub] nswamy commented on issue #10163: Set threads created by MXNetHandler to Daemon threads so they exit when the main thread exits.

nswamy commented on issue #10163: Set threads created by MXNetHandler to Daemon threads so they exit when the main thread exits. URL: https://github.com/apache/incubator-mxnet/pull/10163#issuecomment-374434778 @Roshrini ?

[GitHub] nswamy opened a new pull request #10163: Set threads created by MXNetHandler to Daemon threads so they exit when the main thread exits.

nswamy opened a new pull request #10163: Set threads created by MXNetHandler to Daemon threads so they exit when the main thread exits. URL: https://github.com/apache/incubator-mxnet/pull/10163 ## Description ## Set threads created by MXNetHandler to Daemon threads so they exit when the

[GitHub] nhynes commented on issue #10162: [question] Convert MXNet NDArray to TVM NDArray

nhynes commented on issue #10162: [question] Convert MXNet NDArray to TVM NDArray URL: https://github.com/apache/incubator-mxnet/issues/10162#issuecomment-374432130 If anyone else has this question, there's an example is in the PR, itself:

[GitHub] nhynes closed issue #10162: [question] Convert MXNet NDArray to TVM NDArray

nhynes closed issue #10162: [question] Convert MXNet NDArray to TVM NDArray URL: https://github.com/apache/incubator-mxnet/issues/10162 This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] zhanghang1989 commented on issue #9688: [MXNET-108] Adding BilinearResize2D and AdaptiveAvgPool2d operators

zhanghang1989 commented on issue #9688: [MXNET-108] Adding BilinearResize2D and AdaptiveAvgPool2d operators URL: https://github.com/apache/incubator-mxnet/pull/9688#issuecomment-374431270 Hi @cjolivier01, I have made the changes. Please let me know if you have further comments. Thanks!

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374422617 I removed the network operations and got essentially the same results for master and mxnet-cu90 in

[GitHub] nhynes opened a new issue #10162: [question] Convert MXNet NDArray to TVM NDArray

nhynes opened a new issue #10162: [question] Convert MXNet NDArray to TVM NDArray URL: https://github.com/apache/incubator-mxnet/issues/10162 Is there an easy way to do this or does the TVM bridge need to be expanded? Conversion through numpy is possible but slow. Ideally it would

[GitHub] sxjscience commented on issue #9583: use nd for accuracy calculation

sxjscience commented on issue #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#issuecomment-374425245 Sure. Let’s move the discussion there. Get Outlook for iOS From:

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374422617 I removed the network operations and got essentially the same results for master and mxnet-cu90 in

[GitHub] anirudh2290 commented on issue #10045: Despite so many declarations, Module is still confused by extra names in DataIter

anirudh2290 commented on issue #10045: Despite so many declarations, Module is still confused by extra names in DataIter URL: https://github.com/apache/incubator-mxnet/issues/10045#issuecomment-374423138 This issue seems to be fixed on the master. Can you please check ?

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374422617 I removed the network operations and got essentially the same results for master and mxnet-cu90 in

[GitHub] sandeep-krishnamurthy commented on issue #9078: pip install mxnet dont have warpctc

sandeep-krishnamurthy commented on issue #9078: pip install mxnet dont have warpctc URL: https://github.com/apache/incubator-mxnet/issues/9078#issuecomment-374421744 @hsddlz - please feel free to reopen if you still need support.

[GitHub] sandeep-krishnamurthy closed issue #9078: pip install mxnet dont have warpctc

sandeep-krishnamurthy closed issue #9078: pip install mxnet dont have warpctc URL: https://github.com/apache/incubator-mxnet/issues/9078 This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374411651 I see how that could be a problem, but I think that philosophically MXNet should let the users choose when to block. Typically I

[GitHub] ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation

ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r175615075 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,27 @@ def update(self, labels, preds):

[GitHub] nswamy closed issue #8614: Does mxnet.symbol.SoftmaxOutput support matrix label?

nswamy closed issue #8614: Does mxnet.symbol.SoftmaxOutput support matrix label? URL: https://github.com/apache/incubator-mxnet/issues/8614 This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] nswamy commented on issue #8614: Does mxnet.symbol.SoftmaxOutput support matrix label?

nswamy commented on issue #8614: Does mxnet.symbol.SoftmaxOutput support matrix label? URL: https://github.com/apache/incubator-mxnet/issues/8614#issuecomment-374418062 Closing this, please redirect your question to our user forum

[GitHub] nswamy commented on issue #8591: How do I make a siamese network with pretrained models (esp. keeping the weights the same?)

nswamy commented on issue #8591: How do I make a siamese network with pretrained models (esp. keeping the weights the same?) URL: https://github.com/apache/incubator-mxnet/issues/8591#issuecomment-374417745 @tz-hmc, Hope your question has been answered. For general "how-to" questions,

[GitHub] nswamy closed issue #8591: How do I make a siamese network with pretrained models (esp. keeping the weights the same?)

nswamy closed issue #8591: How do I make a siamese network with pretrained models (esp. keeping the weights the same?) URL: https://github.com/apache/incubator-mxnet/issues/8591 This is an automated message from the Apache

[GitHub] sxjscience commented on a change in pull request #9583: use nd for accuracy calculation

sxjscience commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r175613060 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,27 @@ def update(self, labels, preds):

[GitHub] ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374411651 I see how that could be a problem, but I think that philosophically MXNet should let the users choose when to block. Typically I

[GitHub] ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation

ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r175611091 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,27 @@ def update(self, labels, preds):

[GitHub] ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation

ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r175611091 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,27 @@ def update(self, labels, preds):

[GitHub] ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374411651 I see how that could be a problem, but I think that philosophically MXNet should let the users choose when to block. Typically I

[GitHub] ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374411651 I see how that could be a problem, but I think that philosophically MXNet should let the users choose when to block. Typically I

[GitHub] ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374411651 I see how that could be a problem, but I think that philosophically MXNet should let the users choose when to block. Typically I

[GitHub] ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374411651 I see how that could be a problem, but I think that philosophically MXNet should let the users choose when to block. Typically I

[incubator-mxnet] branch nlp_toolkit updated (086f90a -> 9e4f9d1)

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. discard 086f90a word language model end-to-end example (AWD/RNNModel) (#4) This update removed existing

[GitHub] zheng-da commented on issue #10161: Fix mem alignment.

zheng-da commented on issue #10161: Fix mem alignment. URL: https://github.com/apache/incubator-mxnet/pull/10161#issuecomment-374407010 I didn't know what alignment was required by MKLDNN, so I chose an insanely large alignment. I vaguely remember that 4096 is used by mkldnn for memory

[GitHub] cjolivier01 commented on issue #10161: Fix mem alignment.

cjolivier01 commented on issue #10161: Fix mem alignment. URL: https://github.com/apache/incubator-mxnet/pull/10161#issuecomment-374407219 fair enough :) This is an automated message from the Apache Git Service. To respond to

[GitHub] sxjscience commented on a change in pull request #9583: use nd for accuracy calculation

sxjscience commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r175607170 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,27 @@ def update(self, labels, preds):

[GitHub] sxjscience closed issue #7881: ` list index out of range` and ` NativeOp doesn't support AddTo for output`

sxjscience closed issue #7881: ` list index out of range` and ` NativeOp doesn't support AddTo for output` URL: https://github.com/apache/incubator-mxnet/issues/7881 This is an automated message from the Apache Git Service.

[GitHub] jesterhazy commented on issue #7881: ` list index out of range` and ` NativeOp doesn't support AddTo for output`

jesterhazy commented on issue #7881: ` list index out of range` and ` NativeOp doesn't support AddTo for output` URL: https://github.com/apache/incubator-mxnet/issues/7881#issuecomment-374403806 duplicate of #1983. Try using CustomOp instead.

[GitHub] aaronmarkham closed issue #10159: Scala API docs search is broken

aaronmarkham closed issue #10159: Scala API docs search is broken URL: https://github.com/apache/incubator-mxnet/issues/10159 This is an automated message from the Apache Git Service. To respond to the message, please log on

[GitHub] zhreshold commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

zhreshold commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374401414 @ThomasDelteil Measuring IO with training is not a good idea. Can you pull up with pure IO test?

[GitHub] szha closed pull request #10156: add deprecated profiler_set_state()

szha closed pull request #10156: add deprecated profiler_set_state() URL: https://github.com/apache/incubator-mxnet/pull/10156 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: add deprecated profiler_set_state() (#10156)

This is an automated email from the ASF dual-hosted git repository. zhasheng 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 926f2e2 add deprecated

[GitHub] jesterhazy commented on issue #9078: pip install mxnet dont have warpctc

jesterhazy commented on issue #9078: pip install mxnet dont have warpctc URL: https://github.com/apache/incubator-mxnet/issues/9078#issuecomment-374399607 You can build mxnet from source with the warpctc plugin by following the instructions here:

[GitHub] cjolivier01 commented on issue #10161: Fix mem alignment.

cjolivier01 commented on issue #10161: Fix mem alignment. URL: https://github.com/apache/incubator-mxnet/pull/10161#issuecomment-374399305 Curious why it was set to 4096? This is an automated message from the Apache Git

[GitHub] szha commented on a change in pull request #10160: `gluon.Block.load_params` set the default `ctx` to cpu

szha commented on a change in pull request #10160: `gluon.Block.load_params` set the default `ctx` to cpu URL: https://github.com/apache/incubator-mxnet/pull/10160#discussion_r175601040 ## File path: python/mxnet/gluon/block.py ## @@ -299,7 +300,7 @@ def save_params(self,

[incubator-mxnet] branch nlp_toolkit updated: word language model end-to-end example (AWD/RNNModel) (#4)

This is an automated email from the ASF dual-hosted git repository. zhasheng pushed a commit to branch nlp_toolkit in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/nlp_toolkit by this push: new 086f90a word language

[GitHub] oronanschel commented on issue #8458: Synchronized Batch normalization in multi GPUs training

oronanschel commented on issue #8458: Synchronized Batch normalization in multi GPUs training URL: https://github.com/apache/incubator-mxnet/issues/8458#issuecomment-374395482 Hi, did someone implemented that yet? This is

[GitHub] szha commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

szha commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374394786 The blocking ensures that computation of next batch doesn't start before the computation of the current batch finishes. Without the

[GitHub] ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374392570 Yes that's correct. I see, ideally it would be nice to be able to opt-out of the blocking metrics. What is the rationale behind

[GitHub] leopd commented on issue #10158: Assertion fail when creating very large matrix

leopd commented on issue #10158: Assertion fail when creating very large matrix URL: https://github.com/apache/incubator-mxnet/issues/10158#issuecomment-374391422 Any reasonable workaround until this plan is finished? This

[GitHub] szha commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

szha commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374391456 @ThomasDelteil as I clarified in your comment in the other PR, you probably misunderstood the logic in that PR. I think your comments really

[GitHub] ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

ThomasDelteil commented on issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571#issuecomment-374285733 That seems pretty critical indeed. The previous PR wasn't correct, it was still doing the computation using asnumpy() because it

[GitHub] zheng-da opened a new pull request #10161: Fix mem alignment.

zheng-da opened a new pull request #10161: Fix mem alignment. URL: https://github.com/apache/incubator-mxnet/pull/10161 ## Description ## This is to fix the memory alignment for the memory allocator. When we allocate memory for MKLDNN, we only need to make the memory address aligned to

[GitHub] ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation

ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r175592748 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,27 @@ def update(self, labels, preds):

[GitHub] szha closed issue #6785: Error importing mxnet in python, installed using pip

szha closed issue #6785: Error importing mxnet in python, installed using pip URL: https://github.com/apache/incubator-mxnet/issues/6785 This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation

ThomasDelteil commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r175591965 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,27 @@ def update(self, labels, preds):

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374378588 Thanks @Jerryzcn here is a reproducible example: Play with the `NUM_WORKERS` and `optimized`

[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_r175587494 ## File path: src/operator/adaptive_avg_pooling.cc ## @@ -91,7

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374378588 Thanks @Jerryzcn here is a reproducible example: Play with the `NUM_WORKERS` and `optimized`

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374378588 Thanks @Jerryzcn here is a reproducible example: Play with the `NUM_WORKERS` and `optimized`

[GitHub] cjolivier01 commented on a change in pull request #9688: [MXNET-108] Adding BilinearResize2D and AdaptiveAvgPool2d operators

cjolivier01 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_r175585122 ## File path: src/operator/adaptive_avg_pooling.cc ## @@ -91,7 +94,7

[GitHub] mli opened a new pull request #10160: `gluon.Block.load_params` set the default `ctx` to cpu

mli opened a new pull request #10160: `gluon.Block.load_params` set the default `ctx` to cpu URL: https://github.com/apache/incubator-mxnet/pull/10160 ## Description ## It's convenience to just have ```python net.save_params('net.params') net.load_params('net.params')

[GitHub] cjolivier01 commented on a change in pull request #9688: [MXNET-108] Adding BilinearResize2D and AdaptiveAvgPool2d operators

cjolivier01 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_r175584879 ## File path: src/operator/adaptive_avg_pooling.cc ## @@ -91,7 +94,7

[GitHub] szha commented on a change in pull request #9583: use nd for accuracy calculation

szha commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r175583728 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,27 @@ def update(self, labels, preds):

[GitHub] Jerryzcn commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

Jerryzcn commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374372260 @ThomasDelteil The bug is related to a race condition in memory management, where a space is double

[GitHub] aaronmarkham opened a new issue #10159: Scala API docs search is broken

aaronmarkham opened a new issue #10159: Scala API docs search is broken URL: https://github.com/apache/incubator-mxnet/issues/10159 ## Description The search field does nothing. https://mxnet.incubator.apache.org/api/scala/docs/index.html#package

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374359896 The segfault does not seem to happen with latest master, however latest master does seem to be MUCH

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374359896 The segfault does not seem to happen with latest master, however latest master does seem to be MUCH

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374359896 The segfault does not seem to happen with latest master, however latest master does seem to be MUCH

[GitHub] ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training

ThomasDelteil commented on issue #10042: [MXNET-86] Gluon dataloader crash on speech recognition training URL: https://github.com/apache/incubator-mxnet/issues/10042#issuecomment-374359896 The segfault does not seem to happen with latest master, however latest master does seem to be MUCH

  1   2   3   >