Re: Global Search Now Available on MXNet Website

Awesome work! Thanks a lot for making this desirable feature happen. Lin On Wed, May 20, 2020 at 8:45 AM Yang Shi wrote: > Hi MXNet Community, > > Global search feature is added to the main information pages of MXNet > website. It can search for contents across all site in any version. >

Re: Versioned Dropdown for Python API Docs

Connor, Good job! Thanks for your contribution. The versioned website looks neat and hopefully can reduce a lot of confusion to MXNet users. Best, Lin On Thu, Apr 9, 2020 at 11:03 AM Goggins, Connor wrote: > Hi all, the new production MXNet website (with general API version > dropdown) is

Re: MXNet Bot Demo

Chai, Awesome work. When do we expect this bot to be deployed? Best, Lin On Thu, Mar 12, 2020 at 2:00 PM Chaitanya Bapat wrote: > Hello MXNet community, > > I have built an MXNet Bot that allows PR > Authors, Committers and Jenkins Admins to trigger CI

Re: [apache/incubator-mxnet] [RFC] Apache MXNet 2.0 Roadmap (#16167)

Is there a plan to remove the cudnn_off argument from the neural network operators such as Dropout, Convolution, Pool etc. It creates a few usability issues: (1) Once a model is exported. It requires users to change this flag in all the layers manually if they want to enable/disable cuDNN. When

Re: [VOTE] Release Apache MXNet (incubating) version 1.6.0.rc2

t; > > +1 > > > > Successfully built MXNet 1.6.0rc2 on Linux > > > > Tested for OpPerf utility > > > > For CPU - > > > > https://gist.github.com/ChaiBapchya/d5ecc3e971c5a3c558d672477b4b6b9c > > > > > > > > Works well! > > > > > >

Re: [VOTE] Release Apache MXNet (incubating) version 1.6.0.rc2

+1 Tested Horovod with mnist example. My compiler flags are below: [✔ CUDA, ✔ CUDNN, ✔ NCCL, ✔ CUDA_RTC, ✖ TENSORRT, ✔ CPU_SSE, ✔ CPU_SSE2, ✔ CPU_SSE3, ✔ CPU_SSE4_1, ✔ CPU_SSE4_2, ✖ CPU_SSE4A, ✔ CPU_AVX, ✖ CPU_AVX2, ✔ OPENMP, ✖ SSE, ✔ F16C, ✖ JEMALLOC, ✔ BLAS_OPEN, ✖ BLAS_ATLAS, ✖ BLAS_MKL, ✖

Re: Requesting slack access

Done. Welcome to MXNet community! Lin On Sat, Jan 25, 2020 at 5:48 PM Tajinder Singh wrote: > Hi, > > Please add me to slack work space. My email: tsingh2...@gmail.com > > Thanks, > Tajinder >

Re: [apache/incubator-mxnet] [RFC] Deferred compute in imperative interface to unify imperative and symbolic interface (#16376)

This seems to be a big change to the existing operator mode (imperative and symbolic). Could you please provide more information. AFAIK, symbolic API already does deferred init, imperative API is provided to improve user experience. Based on this RFC, what's the advantage of this new

Re: Request to join slack channel

Done On Wed, Jan 22, 2020 at 12:13 PM Salim Chemlal wrote: > Hi, > > I am an AI adjunct faculty at Old Dominion University and a DL engineer. I > am requesting to be added to mxnet slack channel, email is > drchem...@gmail.com > > Thank you > > Salim >

Re: Slack Access

Invitation sent. Thanks for your interest. Lin On Wed, Jan 22, 2020 at 1:15 AM João Costa wrote: > Hi > > Can you give me access to slack? > > Thanks > João Costa >

[NOTIFICATION] CI Restart

Dear Community, Since Jan 14, 2020, our developers have identified frequently occurrence of test time-out in our CI system. Nick and Pedro have helped to investigate this random test timeout, however, due to the design of CI system the failed instances are already reclaimed and not enough logging

Re: MXNet 1.6 as last release with Python 2 support?

+1 On Fri, Jan 17, 2020 at 10:04 AM Xingjian SHI wrote: > +1. We should move to support Python>=3.5 only. > > Get Outlook for iOS > > From: Lausen, Leonard > Sent: Friday, January 17, 2020 10:02:30 AM > To: d...@mxnet.apache.org >

Re: [DISCUSS] Enforce tighter control on API related changes

on this in the issue #17292, let's > make sure the related details are added there please. > > -sz > > On 2020/01/14 18:24:13, Lin Yuan wrote: > > Hi Sheng, > > > > Thanks for your reply. > > > > 1) Adding a "API Change" label is a good way to f

Re: [DISCUSS] Enforce tighter control on API related changes

of MXNet 2.0, it's a good time to rethink how horovod > integration should work with MXNet. I'm hoping that MXNet 2.0 item 4.11 > AbstractKVStore interface (See #17115) could help simplify and alleviate > the coupling in the current way of integration. > > -sz > > [1] > https://git

[DISCUSS] Enforce tighter control on API related changes

Dear Community, Recently, there were some changes to C APIs that broke another downstream project Horovod: https://github.com/apache/incubator-mxnet/issues/17292. Since we do not have integration tests for downstream project, it becomes critical for us to update APIs with extreme caution. I

Re: Stopping nightly releases to Pypi

t; >>> > >>> On 2020/01/06 18:19:52, "Lausen, Leonard" >>> <mailto:lau...@amazon.com.INVALID>> > >>> wrote: > >>> Consider a user finds a bug in a nightly version but we can't narrow > >>> down the > >

Re: [VOTE] Release Apache MXNet (incubating) version 1.6.0.rc1

We can release one cpu-mkl and one CUDA wheel for testing various applications. Other people can build from source if they want other flavors Lin On Fri, Jan 10, 2020 at 4:00 PM Karan Jariwala wrote: > Yes, agree with your point. But we will be requiring many flavors of pip > wheel. > > MKL/

Re: [VOTE] Release Apache MXNet (incubating) version 1.6.0.rc1

Correction: it was built from source on Ubuntu 16.04 On Tue, Jan 7, 2020 at 11:42 AM Lin Yuan wrote: > +1 > > Build from source on Ubuntu 18 with CUDA/CUDNN/NCCL on and verified it > works with Horovod 0.18.2 > > On Tue, Jan 7, 2020 at 9:55 AM Przemysław Trędak > w

Re: [VOTE] Release Apache MXNet (incubating) version 1.6.0.rc1

+1 Build from source on Ubuntu 18 with CUDA/CUDNN/NCCL on and verified it works with Horovod 0.18.2 On Tue, Jan 7, 2020 at 9:55 AM Przemysław Trędak wrote: > Dear MXNet community, > > This is the vote to release Apache MXNet (incubating) version 1.6.0. > Voting starts today and will close on

Re: Stopping nightly releases to Pypi

+1 for a nightly pip with fixed name. We need this to track mxnet integration with other packages such as Horovod. Sam, when do you think we can have this nightly build with a fixed name? Thanks, Lin On Sun, Jan 5, 2020 at 7:48 PM Skalicky, Sam wrote: > Hi Tao, > > We dont have this yet,

Re: [VOTE] Release Apache MXNet (incubating) version 1.6.0.rc0

ing > these changes? > > -sz > > > On Dec 27, 2019, at 10:30 AM, Lin Yuan wrote: > > > > Are these release blocker? It's very risky to make such last-minute big > > change after code freeze. > > > > Can we do this in the next release? > &

Re: [VOTE] Release Apache MXNet (incubating) version 1.6.0.rc0

g > > > the fix to v1.6.x branch. I will prepare the RC1 once that is ready. > > > > > > Thanks > > > Przemek > > > > > > On 2019/12/20 20:07:36, Lin Yuan wrote: > > > > What's the next step for the release? Should we continue test

Re: Proposal for MXNet website improving

Agree with Patric. We should make performance of MXNet more visible in the website. Lin On Sun, Dec 22, 2019 at 9:43 PM Zhao, Patric wrote: > From my view, performance is a big plus for MXNet and the reason why lots > of people adopted in MXNet. > > I still think we need to have a top-level

Re: [VOTE] Release Apache MXNet (incubating) version 1.6.0.rc0

What's the next step for the release? Should we continue testing this and vote or wait until the https://github.com/apache/incubator-mxnet/issues/17105 is fixed? Thanks! Lin On Wed, Dec 18, 2019 at 12:55 AM Lausen, Leonard wrote: > Thanks Przemysław for managing this release and everyone who

Re: Stopping nightly releases to Pypi

Is there a way to install the latest nightly package without having to specify exact date? Thanks, Lin On Sun, Dec 8, 2019 at 6:13 PM Lausen, Leonard wrote: > From Shanghai, the closest endpoint (automatically chosen endpoint) is in > Tokyo > and download speed for mxnet-mkl was on average

Re: Proposal to make MKLDNN as default CPU backend

Also per Sam's suggestion, we could still release a build without MKLDNN (name it mxnet-nomkldnn?) and track the usage/download for one or two releases. If there is no usage, we could drop that build in the future. Best, Lin On Tue, Nov 19, 2019 at 1:23 PM Lin Yuan wrote: > Just to summar

Re: Proposal to make MKLDNN as default CPU backend

Just to summarize base on the concerns Marco raised and discussed abvove: - AMD CPU (it should work with MKLDNN: https://cwiki.apache.org/confluence/display/MXNET/MXNet+with+Intel+MKL-DNN+-+Performance+Benchmarking ) - ARM CPU (we don't have it today w/o MKLDNN either) - Windows (Windows support

Re: [apache/incubator-mxnet] [RFC] Unified API for Distributed Data Parallel Training (#16795)

In the Limitation, I suppose you meant 'use case 1,3,4', right? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-mxnet/issues/16795#issuecomment-553085374

Re: BytePS-MXNet Integration

Very interesting proposal. I have tried BytePS on some examples and did see better performance than Horovod. I look forward to this integration and feel free to let the community know if any help is needed. Lin

Re: ONNX Support

Hi Anirudh, Could you provide more exact data points regarding the ONNX usage and MXNet version? If no one is actively maintaining ONNX any more, I don't see a compelling reason for an engineer to spend quality time to fix an ONNX test in order for his/her PRs to move forward. Lin On Mon, Oct

Re: Update for 1.5.1 patch release

Ping @Sheng and @Lai who released 1.5.0 for help. Could you please update the Release Process doc after you find the right answers to these questions? Thanks, Lin On Sat, Sep 28, 2019 at 7:49 AM Tao Lv wrote: > Hi dev, > > I'm glad to say that the rc0 of 1.5.1 patch release has passed the

Re: new website, docs code freeze

Looks very neat. Thank you Aaron and many others for launching this! On Fri, Sep 20, 2019 at 7:31 AM Carin Meier wrote: > Nice!!! Congrats everyone! > > On Fri, Sep 20, 2019 at 10:28 AM Aaron Markham > wrote: > > > Alrighty! The new site is launched. You might need to clear your cache. > > > >

Re: [VOTE] Release Apache MXNet (incubating) 1.5.1.rc0

+1 Tested Horovod on GPU On Wed, Sep 18, 2019 at 6:16 AM Zhao, Patric wrote: > +1 > > Tested MKLDNN backend and everything looks great. > > > -Original Message- > > From: Qing Lan > > Sent: Wednesday, September 18, 2019 2:20 AM > > To: dev@mxnet.incubator.apache.org > > Subject: Re:

Re: Code freeze for 1.5.1 patch release

elieve > > this needs to hold up the voting. I believe the voting is only on the > > source. > > > > -Carin > > > > On Mon, Sep 16, 2019 at 7:05 PM Lin Yuan wrote: > > > > > Hi Tao, > > > > > > Thanks for uploading the arti

Re: Code freeze for 1.5.1 patch release

Hi Tao, Thanks for uploading the artifacts. May I know what the current status of Scala, Clojure and R packages and any help you need from the community to complete? Thanks, Lin On Fri, Sep 6, 2019 at 7:35 AM Tao Lv wrote: > Update: > > Artifacts of 1.5.1.rc0 have been uploaded to github and

Re: [Announcement] New Committer - Junru Shao

Congratulations! On Sat, Sep 7, 2019 at 8:14 PM Sheng Zha wrote: > Hi all, > > Please join me in welcoming Junru Shao as a new committer of Apache MXNet > (incubating)! > > Junru made a number of contributions to this project such as cond and > while_loop control-flow > operators, enabling

Re: [Discussion] MXNet 1.5.1 release

gt; > still > >> > > think we can have your fix on the 1.5.x branch? > >> > > > >> > > Progress since last update: > >> > > 1. We received several more proposals in the github thread [1]. I > >> > > humbly > >>

Re: [Discussion] MXNet 1.5.1 release

29, 2019 at 3:23 PM Lin Yuan wrote: > Hi Tao, > > 5) is not a bug. It's just a large tensor support requirement. The PR was > to fix a memory alignment issue introduced in master but not in 1.5.1 > (since you did not cherry pick that PR). So, I have crossed out 5) in the > doc an

Re: [Discussion] MXNet 1.5.1 release

gt; code freeze. > > > > Thanks, > > -tao > > > > [1] https://github.com/apache/incubator-mxnet/issues/15613. > > [2] https://github.com/apache/incubator-mxnet/pull/16026 > > [3] https://github.com/apache/incubator-mxnet/pull/15948 > > [4] https://g

Re: [Discussion] MXNet 1.5.1 release

https://github.com/apache/incubator-mxnet/pull/15762 contains some unrelated changes which is being reverted. Please do not cherry pick it yet. On Mon, Aug 26, 2019 at 4:25 PM Pedro Larroy wrote: > There's a fix that I did which seems to still produce crashes in 1.5 for > some users, which I

Re: [Discuss] MXNet Python 2 Support Deprecation

+1 On Fri, Jul 19, 2019 at 12:03 AM Chaitanya Bapat wrote: > +1 definitely. > > Going forward, > MXNet repo as it stands has ~95,000+ lines of Python code [1] > OpenEdx has a million (10x) LOC and this mammoth effort of porting from > Python 2 to 3 is treated as a separate project named

Re: [VOTE] Release Apache MXNet (incubating) version 1.5.0.rc0

> wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On 2019/06/11 18:53:56, Pedro Larroy < > > pedro.larroy.li...@gmail.com > > > > > > > > > > > > > wrote: > > > >

Re: [VOTE] Release Apache MXNet (incubating) version 1.5.0.rc0

-1 (Tentatively until resolved) I tried to build MXNet 1.5.0 from source and pip install horovod but got the following error: Reproduce: 1) cp make/config.mk . 2) turn on USE_CUDA, USE_CUDNN, USE_NCCL 3) make -j MXNet can build successfully. 4) pip install horovod

Re: [RFC] Support for creation of Large Tensors in MXNet

51 > > > > But as I said before, since we support flatten or reshape > operators, so it's possible for users to convert a tensor with large > element size to a tensor with large dimension size. It possibly will cause > issue there. > > > > To cover more cases, MKL-DNN is going to

Re: [Announcement] New Committer - Yuxi Hu

Congrats Darren! Well deserved. On Fri, May 24, 2019 at 6:27 AM Aaron Markham wrote: > Congrats Darren! > > On Thu, May 23, 2019, 18:48 Zhao, Patric wrote: > > > Congratulations, Darren :) Thanks for your great works in Horovod. > > > > > -Original Message- > > > From: Chaitanya Bapat

Re: [Announcement] New Committer - Kedar Bellare

Welcome on board! Lin On Thu, May 23, 2019 at 9:01 AM Carin Meier wrote: > Please join me in welcoming Kedar Belllare https://github.com/kedarbellare > as > a new committer. > > Kedar has worked on the Clojure package and helped improve it by porting > the Scala image and infer functionality

Re: [DISCUSS] 1.5.0 Release Plan

Hi Lai, One important PR that is currently blocked by a Flaky TensorRT test: https://github.com/apache/incubator-mxnet/pull/15041 I have retriggered it several times. If it fails again, I may need CI team to help disable this test. It has been reported by multiple people:

Re: [Announcement] New Committer - Hao Jin

Congrats! On Tue, Apr 30, 2019 at 11:28 PM Alex Zai wrote: > Congrats Hao! > > On Tue, Apr 30, 2019 at 10:53 PM Steffen Rochel > wrote: > > > congratulation Hao! > > > > On Tue, Apr 30, 2019 at 8:05 AM MiraiWK WKCN wrote: > > > > > Congrats Hao! Welcome! > > > > > >

Re: [Announcement] New Committer - Zhennan Qin

Congrats, Zhennan! Well deserved. Lin On Tue, Apr 30, 2019 at 3:07 PM Zhao, Patric wrote: > Cong, Zhennan. > > Really great works and it makes the MXNet/Quantization flow outstanding > over the world! > > > -Original Message- > > From: Lv, Tao A [mailto:tao.a...@intel.com] > > Sent:

Re: [RFC] Support for creation of Large Tensors in MXNet

Tao, - what's the max size of dimensionality? Which data type is used to define dimensionality (ndims)? We assume the max size of dimensionality is relatively small. Hence `int` data type is used to define ndim - what's the max size of each dimension? Which data type is used to define dimension

Re: [QUESTION] mxnet/Tuple vs nnvm/Tuple

ew locations not changed in the current codebase, we should change them to > use mxnet Tuple as well for better cosmetics. The nnvm/tuple.h can be > deprecated in MXNet. > > On Mon, Apr 15, 2019 at 10:44 PM Lin Yuan wrote: > > > Dear Community, > > > > Curr

[QUESTION] mxnet/Tuple vs nnvm/Tuple

Dear Community, Currently in MXNet there are two Tuple template class defined in mxnet/tuple.h and nnvm/tuple.h respectively. These two templates are higly similar and most part are duplicated except for a couple of functions. However, they were used mixedly in current codebase and causing

Re: Fujitsu Breaks ImageNet Record using MXNet (under 75 sec)

Chai, Thanks for sharing. This is awesome news! Lin On Mon, Apr 8, 2019 at 8:48 AM Chaitanya Bapat wrote: > Greetings! > > Great start to a Monday morning, as I came across this news on Import AI, > an AI newsletter. > > The newsletter talked about Apache MXNet, hence thought of sharing it

[RFC] Higher order gradient support in MXNet

Dear Community, Higher order gradient calculation is required for many applications. However, current MXNet only support higher order gradient for a very limited number of operators. We plan to support the higher order gradient calculation in the autograd package. A design proposal is ready for

Re: [DISCUSS] Rebrand Gluon to MXNet imperative or something MXNet.

; (military > >> > > > connotation). If naming is the only thing we need to figure > out > >> - > >> > > that is a > >> > > > good place to be in __ > >> > > > > >> > > > -- > >>

Re: [DISCUSS] Rebrand Gluon to MXNet imperative or something MXNet.

> > good place to be in __ > > > > > > > > > > -- > > > > > Thanks, > > > > > Denis > > > > > > > > > > On 3/22/19, 4:48 PM, "Mu Li" wrote: > > > >

Re: [DISCUSS] Rebrand Gluon to MXNet imperative or something MXNet.

@Junru Thanks for the clarification. Given that we already have courseware and books with Gluon, it makes sense to brand “Mxnet Gluon” with Gluon being the high level API of mxnet @Tianqi what’s the roadmap of GluonNLP/GluonCV? Are they positioned to be high level API of MXnet or some

Re: [DISCUSS] Rebrand Gluon to MXNet imperative or something MXNet.

+1. Just to give some of my real experience: 1) I advertised a recent GluonNLP blog and many responses are "This seems nice. So is Gluon a new library to replace MXNet?" 2) We visited customers in a unicorn company who showed interests in MXNet but none of the engineers knew the relationship

Re: [Announcement] New Committer - Patric Zhao

Congrats, Patric! On Thu, Mar 21, 2019 at 10:32 AM Yuxi Hu wrote: > Congrats, Patric! Well deserved! > > On Wed, Mar 20, 2019 at 1:08 PM kellen sunderland < > kellen.sunderl...@gmail.com> wrote: > > > Congrats Patric! > > > > On Sun, Mar 17, 2019 at 10:34 PM Hagay Lupesko > wrote: > > > > >

Re: [DISCUSS] Process to remove deprecated operators

share some immature ideas about deprecating operators. Not > only adopting semantic versioning, but also should we provide enough > informative error message for customers to understand how to replace > deprecated operators with new ones. > > Thanks, > Junru > > On W

Re: [DISCUSS] Process to remove deprecated operators

elease. > > -sz > > On Wed, Feb 27, 2019 at 8:53 PM Lin Yuan wrote: > > > Dear Community, > > > > In MXNet there are many legacy operators such as this > > < > > > http://mxnet.incubator.apache.org/versions/master/api/python/symbol/symbol.

[DISCUSS] Process to remove deprecated operators

Dear Community, In MXNet there are many legacy operators such as this that has been marked DEPRECATE for several releases. However, these operators still exist in

Re: [Announce] Runtime feature detection

Thanks, Pedro for contributing this long awaiting feature. I can immediately use it for Horovod project now. Bravo! Lin On Tue, Feb 12, 2019 at 2:42 AM Pedro Larroy wrote: > An update on this topic, Sheng just merged the refinements to the > feature detection so it's now a single API call. (

Re: [RESTARTING][VOTE] Release Apache MXNet (incubating) version 1.4.0.rc2

+1 binding Horovod is going to release it's 0.16.0 in the coming week with MXNet integration. We need to release 1.4.0 which includes all the dependencies for Horovod integration. Best, Lin On Mon, Feb 11, 2019 at 9:30 PM Steffen Rochel wrote: > Dear community - > based on Justin's and

Re: [Announcement] New Committer -- Steffen Rochel

Welcome Steffen! Lin On Mon, Feb 4, 2019 at 7:53 PM kellen sunderland < kellen.sunderl...@gmail.com> wrote: > Great news. Congrats Steffen. > > On Mon, Feb 4, 2019, 5:29 PM Thomas DELTEIL wrote: > > > Welcome Steffen! > > > > On Mon, Feb 4, 2019, 15:55 Marco de Abreu wrote: > > > > >

Re: [VOTE] Release Apache MXNet (incubating) version 1.4.0.rc2

+1 build from source on MacOS 10.13.6 and tested mxnet-to-coreml converter. On Mon, Feb 4, 2019 at 9:03 AM Indhu wrote: > +1 > > Build from source and tested few examples from the examples folder. > > Thanks, > Indu > > > > On Fri, Feb 1, 2019 at 6:21 PM Steffen Rochel > wrote: > > > Hi Sheng

Re: [Announcement] New Committer -- Lin Yuan

gt;> > >> Dear Community: > >> > >> Please join me to welcome Lin Yuan(@apeforest) as a new committer of > >> Apache(incubating) MXNet! > >> > >> He has contributed to various improvements, including better > compatibility > >> of larg

Re: Horovod-MXNet Integration

/mpi-operator) lately and it > would > be interesting to see an example of using Horovod + MXNet + Kubeflow using > MPI Operator. Feel free to reach out (@terrytangyuan > <https://github.com/terrytangyuan>) if you encounter any issues. > > Best, > Yuan > > > On

Re: Apache MXNet v1.4.0 release status

Hi Steffen, I would like to ask to include one more PR for 1.4.0.rc1: https://github.com/apache/incubator-mxnet/pull/13845 This PR exports exception handling API of MXNet. It is needed by Horovod with MXNet integration to elegantly throw exception at Python level rather than a C++ abort.

Re: [Annoucement] New Committer -- Iblis Lin

Welcome Iblis, Great to see a good Julia support in MXNet! Lin On Sat, Jan 5, 2019 at 12:32 PM Marco de Abreu wrote: > Welcome Iblis, > > great to have you on board! > > -Marco > > Am Sa., 5. Jan. 2019, 21:13 hat Carin Meier > geschrieben: > > > Please join me in welcoming Iblis Lin as a new

Re: [Question] UI change policy in MXNet

://github.com/apache/incubator-mxnet/pull/5499 . To summarize: > contrib > was created for ops for which we provide limited guarantees with respect to > backward compatibility, interface changes, testing etc. > > Anirudh > > On Thu, Dec 20, 2018 at 1:00 PM Lin Yuan wrot

[Question] UI change policy in MXNet

Dear Community, As a contributor, I would like to know the current policy for updating UI of an operator. I understand UI change should be introduced in major release not minor release. However, it is still not quite clear to me regarding the UI change process: 1) Which guideline should we

Re: [Annoucement] New Committer -- Da Zheng

Congrats! On Mon, Dec 17, 2018 at 9:19 AM Steffen Rochel wrote: > Congratulation Da! > > On Mon, Dec 17, 2018 at 9:02 AM Tianqi Chen wrote: > > > Dear Community: > > > > Please join me to welcome Da Zheng as a new committer of the MXNet. > > > > Da is the main author of MKL-DNN integration and

[PROPOSAL] Large tensor support in MXNet

Dear Community, As some of you may have already encountered, MXNet has a limitation in supporting tensors of size greater than ~4.3 billion elements (2^32). The root cause is because in MXNet backend 32-bit integer type is used as the default integer data type for both computation and storage in

Re: [Announce] Upcoming Apache MXNet (incubating) 1.4.0 release

Hi Steffen, Can we add the following PR to 1.4.0 release: https://github.com/apache/incubator-mxnet/pull/13452 It's just a Python API returning header path so it should not cause any regression issues. But it is required for Horovod to integrate MXNet. It's better to have this in a minor

Re: [Announce] Upcoming Apache MXNet (incubating) 1.4.0 release

https://github.com/apache/incubator-mxnet/pull/13452 is needed in 1.4.0 to support Horovod integration project. Thanks! Lin On Thu, Nov 29, 2018 at 1:40 PM Davydenko, Denis < dzianis.davydze...@gmail.com> wrote: > I suggest to include this issue into tracked ones for the release: >

Re: [Question] Difference between "Feature" and "Feature request" labels in Github

22a5732e61f6fba1f89ac8@%3Cdev.mxnet.apache.org%3E > > > > We need someone with committer privileges to fix it. > > > > > > Thanks > > Anirudh > > > > > > > > On Tue, Nov 13, 2018 at 4:36 PM Lin Yuan wrote: > > > >

[Question] Difference between "Feature" and "Feature request" labels in Github

Dear Community, I often see there are "Feature" and "Feature request" labels in Github issues. May I know the difference? If they are meant to be the same thing, can we only keep one of them? Thanks, Lin

Catch divide-by-zero floating number exception in backend

Dear MXNet Community, I recently found the NaN errors sometimes could be due to some divide-by-zero float number bugs in engine backend. However, by default, such an exception will not be thrown. I added a signal trap to catch this error (https://github.com/apache/incubator-mxnet/pull/13190) and

Re: [Announce] Upcoming Apache MXNet (incubating) 1.3.1 patch release

Hi Anton, Thanks for helping the release. The following PRs are needed by customers who want to use deterministic CUDNN convolution algorithms: https://github.com/apache/incubator-mxnet/pull/12992 https://github.com/apache/incubator-mxnet/pull/13049 Thanks！ Lin On Tue, Nov 6, 2018 at 1:51 PM

Re: [DISCUSS] Speedup non-code PR in CI

Kellen and Pedro, Thanks for your pointers. I am not an expert in CI but one naive speedup I can see is that if the PR only contains *.md file, then skip the build and testing cycles. This can make documentation/correction easier and save computation resource for other needed tests. Any side

[DISCUSS] Speedup non-code PR in CI

Dear Community, I recently submitted a few small PRs with only changes in README files. However, I noticed they still triggered the full cycle of CI including build and test on all platforms. Do we have a plan to speed up this process, maybe skipping non-code related PRs in CI? Sorry, if this

Re: Horovod-MXNet Integration

Hi Mu, Darren (@yuxihu ) and I have been working on releasing MXNet-Horovod integration in production. We have made some changes on both MXNet and Horovod sides. The changes on MXNet side have mostly been merged and we are working to merge code to horovod repo. We will

Re: [Discussion] Recognise Reviewers, Besides Committers and PMC

+1 sounds like a great idea. We also need a mechanism to identify “good reviewers”. Maybe we can count the number of :thumsup: in each review. Or any other better way? On Fri, Oct 19, 2018 at 8:22 PM Tianqi Chen wrote: > Dear MXNet Community: > > There is a great discussion going on in terms of

Re: CUDNN algorithm selection failure

elated to resource usage / allocation / leak in the worst case. > > Could be that there was not enough memory GPU memory at the time of test > execution. But I'm just speculating, hence my original question. > > Pedro. > > On Mon, Oct 1, 2018 at 8:16 PM Lin Yuan wrote: > &g

Re: [DISCUSS] Use modernized C++11 range loops uniformly throughout the project

+1 Using range-based for-loop whenever possible improves code readability and makes code less prone to human error. I did some preliminary research on Google and did not find any complaint about its performance drawback. Here is one piece from StackOverflow for reference:

Re: [LAZY VOTE] Consolidating developer guide in one place (cwiki preferred)

of these > epics. > > Cheers, > Aaron > > On Wed, Sep 26, 2018, 00:02 Lin Yuan wrote: > > > Hi Aaron, > > > > Do we have a resolution for this proposal yet? Recently, there have been > > many asks for a better documentation for MXNet developers. I thin

Re: [LAZY VOTE] Consolidating developer guide in one place (cwiki preferred)

at 1:55 PM Lin Yuan wrote: > +1 > > On Tue, Sep 4, 2018 at 1:46 PM Aaron Markham > wrote: > >> I'd like to call for a lazy vote on this before proceeding. Already had >> some +1s but let's be sure. >> >> The vote is to move developer guide inf

Re: [DISCUSS] Build OSX builds in CI (possibly with TravisCI).

gt; > > > > > > Tianqi > > > > > > > > > On Wed, Sep 5, 2018 at 9:04 AM kellen sunderland < > > > > > > > > > kellen.sunderl...@gmail.com> wrote: > > > > > > > > > > >

Re: [DISCUSS] Build OSX builds in CI (possibly with TravisCI).

Hi Kellen, Many thanks for your and Marco's effort! I think this is a very crucial piece to improve MXNet stability. To add some data points: 1) Customers using CoreML to MXNet converter were blocked for a while because the converter was broken and no unit test was in place to detect that. 2)

Re: [LAZY VOTE] Consolidating developer guide in one place (cwiki preferred)

(cwiki). > > > > > > Note: Permissions on cwiki is currently not well managed with many people > > having full admin rights to edit/create/delete pages. Should be fine for > > now, but, when we start accumulating many documents and resources, we > > should prob

Re: Propose to discontinue supporting Apache MXNet on Windows 7

in the > > > > > sense > > > > > > > that it would be easy to maintain for mxnet devs and allow > us > > > to use > > > > > > modern > > > > > > > cpp on older windows machines without using vs 2013(which I > >

Re: Propose to discontinue supporting Apache MXNet on Windows 7

< > > > > kellen.sunderl...@gmail.com> wrote: > > > > > > > > > +1 (non-binding) > > > > > > > > > > On Wed, Aug 29, 2018, 1:18 AM Anirudh Acharya < > > > anirudhk...@gmail.com> > > > > > wro

Propose to discontinue supporting Apache MXNet on Windows 7

Dear Community, Currently, our MXNet installation guide for Windows does not work for Windows 7. e.g. Microsoft Visual Studio 2015 is not supported on Windows 7 . In addition,

Re: build from source instructions

being more confusing than helpful. It's > > difficult to debug any issues when the entire instruction is to run a > > single script. - Sina > > > > > > > > On 8/28/18, 9:46 AM, "Lin Yuan" wrote: > > > > Aaron, > > > >

Re: Updating MXNet's Cub

+1 On Tue, Aug 28, 2018 at 12:39 AM Hagay Lupesko wrote: > Thanks for the feedback Chris. Will follow up. > > On Fri, Aug 24, 2018 at 10:53 AM Chris Olivier > wrote: > > > +1 for pointing to NVidia's repo for the newer Cub and subsequent > versions. > > > > On Fri, Aug 24, 2018 at 10:01 AM

Re: build from source instructions

Aaron, I agree the installation page is very confusing to me. When I first tried to build MXNet from source on MacOS, I was totally confused about the instruction. Why was it vastly different from building from source on Linux given these two OS have similar shell commands. I feel the automatic

Re: Consolidating developer guide in one place (cwiki preferred)

developer-focused content be > > placed on the wiki. > > > > On Tue, Aug 14, 2018 at 10:40 AM, Lin Yuan wrote: > > > > > Dear MXNet community, > > > > > > As a developer, I noticed we have some developer guide scattered in > > > different websites

Consolidating developer guide in one place (cwiki preferred)

Dear MXNet community, As a developer, I noticed we have some developer guide scattered in different websites (mxnet.io, cwiki): E.g. How to Create New Operators (Layers): [ https://mxnet.incubator.apache.org/faq/new_op.html] A Guide to Implementing Sparse Operators in MXNet Backend [

Enabling shared filter in JIRA

Dear MXNet Community, As we are trying to create our Scrum board on JIRA, I noticed that we do not have the permission to create shared filter, even as an administrator. This has limited us to create scrum boards for different components of the project. I will really appreciate if someone in the

  1   2   >