Re: new website, docs code freeze

/google_analytics.js On Tue, Sep 24, 2019 at 4:23 PM Mu Li wrote: > I think the reason is that the google tracker is not included in the new > website. > > On Tue, Sep 24, 2019 at 4:17 PM Marco de Abreu > wrote: > >> Hello, >> >> I checked the Google Analytics

Re: new website, docs code freeze

I think the reason is that the google tracker is not included in the new website. On Tue, Sep 24, 2019 at 4:17 PM Marco de Abreu wrote: > Hello, > > I checked the Google Analytics statistics and the launch of the new > website reduced the traffic by over 80%: > > [image: image.png] > > (Please

Re: Disabling, circumventing and altering CI checks

Hi Marco, Thank you for helping CI all the time. You did an incredible job on it. Please let me explain why it’s urgent we need to update our CI to allow fast developing. In this summer, we managed to hire a large amount of interns to help, they did great to contribute to MXNet. But CI is

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

> > I am disappointed when I see articles that use Gluon, but don't mention > MXNet [1]. The only sentence contains Gluon is "Our model implementations and training scripts are publicly available in GluonCV ", not referring Gluon as a framework. But I agree that this paper should mention MXNet.

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

-aws-and-microsoft/ On Thu, Mar 28, 2019 at 2:04 PM Isabel Drost-Fromm wrote: > > > Am 28. März 2019 21:53:16 MEZ schrieb Mu Li : > > > >The reason why we call it GluonCV instead of MXNetCV is because MXNet > >is a > >trademark owned by Apache, while Gluon d

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

Hi Lieven, The reason why we call it GluonCV instead of MXNetCV is because MXNet is a trademark owned by Apache, while Gluon doesn't have this issue. Best Mu On Sat, Mar 23, 2019 at 1:52 AM Lieven Govaerts wrote: > Hi, > > On Sat, 23 Mar 2019 at 00:02, Pedro Larroy > wrote: > > > Hi dev@ > >

Re: Gluon fit API- Design proposal

It's great to see the proposal has a list of models that this API should cover. But note that the D2L book has a very simplified train function (I wrote it). It's oversimplified compared to what we are using in real life. Kaggle competition solutions and popular github projects are closer to what

Re: MXNet Community Monthly Updates

> > Digest > > > > > for MXNet (weekly newsletter) - which is on similar lines (can be > > made > > > > into > > > > > Monthly if it sounds good). > > > > > > > > > > Here's the quip doc - > > > &g

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

ds for MXNet-Modified (military > connotation). If naming is the only thing we need to figure out - that is a > good place to be in __ > > -- > Thanks, > Denis > > On 3/22/19, 4:48 PM, "Mu Li" wrote: > > Gluon is about imperative neural network training

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

Gluon is about imperative neural network training and data loading. ndarray is another large imperative module. Besides, Gluon also supports symbolic execution after hybridizing. mxnet imperative might not be a good name for it. Another choice is high-level API, that's how TF talks about Keras.

Re: CI unstable

I saw CI is downloading from data.dmlc.ml. Changing it data.mxnet.io should fix this issue. Say http://data.dmlc.ml/models/imagenet/inception-bn/Inception-BN-0126.params -> http://data.mxnet.io/models/imagenet/inception-bn/Inception-BN-0126.params On Thu, Mar 21, 2019 at 11:57 AM Anton Chernov

MXNet Community Monthly Updates

Dear Community, I propose to send a monthly summary to users to broadcast the recent progresses in the community. It will not only include new features added into MXNet, but also various community activities. Here is an example: Tutorials - 10 new lectures teaching at UC Berkeley - Video record

Re: direction for documentation across various APIs that share common doc source

The original design is putting psudo-code in cc files (e.g. ndarray.cc ) that are languange indepent, then having python codes in .py files (e.g. ndarray_doc.py

change the forwarding of gluon-crash-course.mxnet.io

Dear community, I'm changing the forwarding address of gluon-crash-course.mxnet.io to https://beta.mxnet.io/guide/crash-course/index.html. The http version should be effective in a few minutes, while the https version may take up to 24 hours. It means https://gluon-crash-course.mxnet.io/ might be

Redesign of the MXNet document site

Dear Community, I have been working with several Amazon folks in the past few weeks for a major update of the document website. Now we have a minimal demo at https://beta.mxnet.io/ and would like to ask for feedback. What's new: 1. A new theme based on the material design that looks modern 2.

Re: Horovod-MXNet Integration

Thanks for your contribution, Carl. I remember I left a comment on the proposal, but today I found it was disappeared. My suggestion is trying best to not change the existing API. The reason is that we need to change all trainers on the frontend that uses the existing kvstore APIs, which may

Re: Propose to discontinue supporting Apache MXNet on Windows 7

According to google analytics, ~12% users who visited mxnet's website are using Windows 7. It's a significant number. Even though we cannot conclude that all of these users will run MXNet on Windows 7, I suggest we still support win7. BTW, anyone who can access mxnet's google analytics report can

Re: Help with understanding docs

On Wed, Aug 8, 2018 at 2:20 PM, Isabel Drost-Fromm wrote: > > > Am 8. August 2018 21:34:47 MESZ schrieb Tianqi Chen < > tqc...@cs.washington.edu>: > >As far as I understand, these are packages that are build on top of > >gluon > >and runs on MXNet, > > Going through the hosted subdomains that

Re: Help with understanding docs

MXNet's website is http://mxnet.incubator.apache.org/. I own mxnet.io, which was mxnet's main site before, but now we are using its subdomains to host various packages in the mxnet ecosystem. On Wed, Aug 8, 2018 at 12:13 PM, Isabel Drost-Fromm wrote: > Hi, > > I had a bit of time recently to

Re: Publish MXNet images to DockerHub

cu and mxnet-cumkl are not > actively maintained. > > > > - > > Anirudh > > > > > > On Sat, Jul 21, 2018 at 9:09 PM Mu Li wrote: > > > Surprisingly only the python binding is actively maintained. I remember > we > > can easily push all bindings into docker h

Re: Publish MXNet images to DockerHub

Surprisingly only the python binding is actively maintained. I remember we can easily push all bindings into docker hub through the script in https://github.com/apache/incubator-mxnet/tree/master/docker. On Sat, Jul 21, 2018 at 5:03 PM, Anirudh Acharya wrote: > Hi, > > Docker Hub(

Re: Remove Caffe functionality

7/19/18, 1:27 PM, "Steffen Rochel" wrote: > > Hi Mu - do we have any indication how many users we do have on the > functionality? I see only two github issues from 2017 open. > > On Thu, Jul 19, 2018 at 12:54 PM Mu Li wrote: > > > Hi Anton, >

Re: Remove Caffe functionality

Hi Anton, It's understandable that Caffe is old and its community is shrinking. For this very reasons, existing caffe users are looking for alternatives. The current converter is important for such users to transit to MXNet. It addresses an important issue that: how I can reuse my previous caffe

Re: Request for feedback: proposal for MXNet SDK Office hours

A minor suggestion: rename MXNet SDK to AWS MXNet SDK or Amazon MXNet SDK. On Wed, Jul 18, 2018 at 2:22 PM, Davydenko, Denis < dzianis.davydze...@gmail.com> wrote: > Hello, MXNet community, > > Following up on recent announcement of office hours introduction from > MXNet Berlin team, we are

Re: [VOTE] Release MXNet version 1.2.1.RC0 (Patch Release)

0 to save models be able to load parameters in > >> 1.2.1 using the reverted load_params? (i.e. is it backward compatible) > >> > >> -sz > >> > >> > >>> On Mon, Jun 25, 2018 at 4:07 PM, Anirudh > wrote: > >>> >

Re: [VOTE] Release MXNet version 1.2.1.RC0 (Patch Release)

v1.2.1 will print a deprecating warning message when calling save_params. We should tell users clearly to replace "save_params" with "save_parameters" or something else. On Mon, Jun 18, 2018 at 6:52 PM, Anirudh wrote: > Hi, > > This is the vote to release Apache MXNet (incubating) version

Re: users@mxnet

I like the idea to have a user mail list, it should be able to cover the user segment who prefer to use mail lists. Though my concern is how many of us will be volunteered to maintain this list. We have several contributors actively maintain the forum, including Sina, Thom, and Thomas. We need

Re: Reverting pull request

forcing > > them. > > > > > > I could've also made a script that prevents any PR from being > > self-merged, > > > but I thought our committers are responsible enough to follow our own > > rules > > > without systems actually enforcing them.

Re: Reverting pull request

's a fix? > > If we keep going backwards like this, no progress can be made. > > > > On Fri, Jun 15, 2018 at 2:37 PM, Mu Li wrote: > > > > > Agree that major changes need more extensive reviews. But we cannot > > ignore > > > that both reviews and CI c

Re: Reverting pull request

actually enforcing them. I won't waste my time working on > that script, but from now on I will revert every single PR (except > emergency cases) that has been self-merged without approval. > > -Marco > > On Fri, Jun 15, 2018 at 2:15 PM Mu Li wrote: > > > Why revert

Re: Reverting pull request

Why reverting instead of fixing the bugs? Static memory aims to reduce memory allocation, it's a key feature to bridge the perf gap between gluon and symbol. On Fri, Jun 15, 2018 at 2:06 PM, Marco de Abreu < marco.g.ab...@googlemail.com.invalid> wrote: > Hello, > > I'm reverting

Re: landing pages for the website

Thanks for the proposal, Aaron. The landing page idea is great. I feel we should group MXNet topics into landing pages, such as Gluon you mentioned, also performance, and deployment. But I'm less sure if we want to add too many contents from 3rd party projects into MXNet, otherwise, it may make

Re: [GitHub] szha closed pull request #11154: Revert "[MXNET-503] Website landing page for MMS (#11037)"

Hi Sheng, I suggest to put down a reason for such actions later. It may confuse other contributors, e.g., Steffen raised his concern in a private thread. Best Mu > On Jun 10, 2018, at 7:42 PM, Sheng Zha wrote: > > Thanks, Henri. I was reverting the commit on a PR that another committer >

Re: Accidental commits to 1.2.0 branch using Maven apache-release profile

ull requests for the 1.2 branch which are not from key > contributors. I'm also aware of a few Amazon internal customers who are > currently actively working off the 1.2 branch to prepare for 1.2.1 release, > so I'd rather err on the side of caution here. > > -marco > > Mu Li schr

Re: Accidental commits to 1.2.0 branch using Maven apache-release profile

fits >>>>> >>>>> -Marco >>>>> >>>>> Naveen Swamy schrieb am Sa., 9. Juni 2018, >> 21:39: >>>>> >>>>>> Hi Mu, >>>>>> >>>>>> No, it isn't necessary to have those commits. I d

Re: Accidental commits to 1.2.0 branch using Maven apache-release profile

Hi Naveen, Thanks for the clarification. Is there necessary to keep these commits submitted by the maveen plugin in the repo? Otherwise, can we squash these commits and force push it to a single commit? It's good to have a stable release only patched with meaningful commits. Best, Mu > On

Re: Vote to stop using JIRA

Hi Naveen, Can we clarify that how JIRA improved the project? One key Apache spirit is about transparent, but I didn't see the current way we are using JIRA improves it. Comparing to Spark, one major problem is that we have 10x less active contributors. Even PyTorch is 3x more active (#commits,

Re: data.mxnet.io is gone (correction)

Up now. (The EBS volume was unmounted.) I'll move all data into S3 On Fri, Apr 13, 2018 at 11:51 AM, Aaron Markham wrote: > So anyone know what happened? It's a critical resource for datasets > and examples and more... and all the files seem gone. > >

Re: MXNet Name Change?

Agree that MXNet, the combination of Minerva and CXXNet, which can be interpreted as mixed-net, is hard to be pronounced. But rebranding a name is a very big decision. We need a very carefully designed marketing plan for it. A choice is that we can gradually refer MXNet as a backend, and talk

Re: Unit tests removed

> Per Apache: > > "A code-modification proposal may be stopped dead in its tracks by a -1 > vote by a qualified voter. This constitutes a veto, and it cannot be > overruled nor overridden by anyone. Vetos stand until and unless withdrawn > by their casters." > > > >

Re: Unit tests removed

y on someone adding the tests back later, that’s > just a good intention. > > Hen > > On Thu, Feb 1, 2018 at 11:11 Mu Li <muli@gmail.com> wrote: > > > I think simply reverting the MKL PR that Da has been working for a half > > year and immediately enabling the c

Re: Unit tests removed

zing). > > -Kellen > > On Thu, Feb 1, 2018 at 8:11 PM, Mu Li <muli@gmail.com> wrote: > > > I think simply reverting the MKL PR that Da has been working for a half > > year and immediately enabling the cpp tests in the CI without reaching an > > agreement

Re: Unit tests removed

I think simply reverting the MKL PR that Da has been working for a half year and immediately enabling the cpp tests in the CI without reaching an agreement with Da seriously hurts his feelings. I put the details at the end of https://github.com/apache/incubator-mxnet/pull/9661 On Thu, Feb 1,

Re: Module maintainers proposal

on't aware any other contributor use the direct email way. So I didn't clarify it on the proposal. On Tue, Jan 9, 2018 at 11:47 AM, Isabel Drost-Fromm <isa...@apache.org> wrote: > > > Am 9. Januar 2018 18:25:50 MEZ schrieb Mu Li <muli@gmail.com>: > >We should encoura

Re: Module maintainers proposal

whole. > >You > >probably meant maintainer instead of owner. > > > >Best, > >Sebastian > > > >On 09.01.2018 08:24, YiZhi Liu wrote: > >> +1 for adding @CodingCat (Nan Zhu) as owner of Scala package. > >> > >> 2018-01-08 21:

Module maintainers proposal

It has been a while for discussing having maintainers for individual modules. A maintainer for a module will be the main person who reviews and approves PRs contributed to that module. Currently, some maintainers are listed in the file CODEOWNERS

Re: Increase indentation limit from 100 to 120 characters

I read more google codes instead of amazon codes. That's why I used Google for example, not because Google is "our competitor", which term is actually disallowed to say by amazon employees. I think the question is not "there's a good reason *not to* increase", but "why we want to make such a

Re: Increase indentation limit from 100 to 120 characters

I think this code should be refactored instead of increasing the indent. Puting the test codes within 7 nested for loops is hard to read. Maybe we can split it into multiple functions, or create some templates, or try "using mxnet::op::BatchNormV1Prop" In general, I feel 80-column is not a big

Re: request for reviewing PR in ps-lite

@yizhi, please go ahead. I'm on NIPS this week, probably don't have enough time to dive deep into the codes On Tue, Dec 5, 2017 at 1:51 PM, YiZhi Liu wrote: > to my understanding, this is not about fault-tolerance, i.e., restart when > worker/server fail, right? > > I can

Re: Protected master needs to be turned off

> > Tianqi > > > On Fri, Dec 1, 2017 at 9:41 AM Mu Li <muli@gmail.com> wrote: > > > Totally agree that CI is useful. Actually, I wrote the jenkinsfile and > > setup the jenkins server before moving to apache server. I just mention > > that we cannot rely on

Re: Protected master needs to be turned off

Totally agree that CI is useful. Actually, I wrote the jenkinsfile and setup the jenkins server before moving to apache server. I just mention that we cannot rely on the CI test. It currently covers operator unittests and regression test on several cnns. But the code coverage isn't great. If a PR

Re: Protected master needs to be turned off

I barely remember CI caught bugs except for lint, but it definitely slows down the code merge. I understand the general concerns for removing master protection. So I propose to use the dev branch to merge changes until the CI is stable. And make the nightly build build the dev branch instead

Re: AWS contributing ONNX-MXNet

Thanks for pointing it out, Hagay. I actually missed the "Special thanks to dmlc/nnvm team" sentence. I was looking for Zhi and Tianqi's names, because the onnx converter was mainly done by both of them. However I do not entirely agree that I "had no comments". I actually pointed out the initial

Re: AWS contributing ONNX-MXNet

Thanks for letting the mxnet community know it. Given this new converter is derived from the nnvm-onnx converted developed by Zhi and Tianqi, see this link and that link

Re: [Proposal] Stabilizing Apache MXNet CI build system

g. It's also compatible with Jenkins. > > On Fri, Oct 20, 2017 at 1:51 PM, Tianqi Chen <tqc...@cs.washington.edu> > wrote: > > > +1 > > > > Tianqi > > On Fri, Oct 20, 2017 at 1:39 PM Mu Li <muli@gmail.com> wrote: > > > > > +1 &

Re: New Apache MXNet logo idea

Clearly better than the previous logos we had. One concern is that the logo color is inconsistent with the current doc theme: https://mxnet.incubator.apache.org/versions/master/index.html. We need to change either the logo or the doc theme. On Fri, Oct 20, 2017 at 1:02 PM, sandeep krishnamurthy <

Re: Request for suggestions- Supporting onnx in mxnet

t; As for where the code should sit, we have seen onnx's support > for > > > > > caffe2 > > > > > > > sitting on a separate repo. My suggestion would be put code > > under > > > > > > nnvm/top > >

Re: Request for suggestions- Supporting onnx in mxnet

> > > and > > > > > influence the model exchange format). > > > > > > > > > > I am glad to see the discussion happening and surely there is > doubt, > > as > > > > > with every big step of changes. But with the rapid

Re: Request for suggestions- Supporting onnx in mxnet

d surely there is doubt, > as > > > > with every big step of changes. But with the rapidly changing pace > of > > > deep > > > > learning systems, this is the direction that we thought is most > > > promising. > > > > We can c

Re: Request for suggestions- Supporting onnx in mxnet

I'm speaking under my "MXNet contributor" hat. It will be sad that our new model format and compiler is not supported by our own contributors. It puts us in a bad position to reach out to outside to ask for support. If you really what to do it with the onnx <-> mxnet way, I suggest putting the

Re: Request for suggestions- Supporting onnx in mxnet

Hi Hagay, As mentioned on my previous thread, " MXNet has lots of users already using the Symbolic API which hopefully mean that is a mature API that is not likely to have breaking changes or major issues." actually indicate NNVM is stable. Because MXNet uses NNVM's symbolic.h directly, see

Re: Request for suggestions- Supporting onnx in mxnet

I cannot get the point. MXNet relies on NNVM. In fact, the Symbol object in MXNet is defined on NNVM. On Wed, Oct 18, 2017 at 6:09 PM, Chris Olivier wrote: > My $0.02: > > NNVM is not currently an Apache module. It’s under dmlc umbrella, whose > direction and governance

Re: proposal to shut down apache-mxnet.slack.com

e private channels to the-asf? > Steffen > > > On Mon, Oct 16, 2017 at 10:34 PM Mu Li <muli@gmail.com> wrote: > > > there are several private channels on apache-mxnet.slack.com, how we are > > going to deal with it? > > > > On Mon, Oct 16, 2017 at

Re: proposal to shut down apache-mxnet.slack.com

there are several private channels on apache-mxnet.slack.com, how we are going to deal with it? On Mon, Oct 16, 2017 at 10:19 PM, Hen wrote: > +1. > > On Mon, Oct 16, 2017 at 10:01 PM, Chris Olivier > wrote: > > > +1. Nuke it. > > > > > > On Mon, Oct

Re: MXNet project web page

I suggest using the current logo on http://mxnet.io We on the way to transfer the website to http://mxnet.incubator.apache.org On Fri, Aug 4, 2017 at 2:35 AM, Isabel Drost-Fromm wrote: > Hi, > > I'm currently in the process of creating a slide deck and wanted to use the >

Re: IMPORTANT: Link accounts to Apache today 7/12

Done, my apache id is muli On Wed, Jul 12, 2017 at 3:02 PM, Ly Nguyen wrote: > As mentioned in the migration FAQ, if you currently have any kind of write > access to MXNet repo, once migration happens, YOU WILL LOSE IT unless you > do the following. > > If you haven't

Re: ZeroMQ licensing in Apache MXNet

> > > > I would strongly encourage podlings to try to leverage what the ASF > > > provides, we ship a number of messaging systems that may be better > from a > > > licensing stand point - ActiveMQ, RocketMQ, Pulsar. > > > > > > John > > > > &

Re: Licensing update per code move

Created a PR to resolve this issue. https://github.com/dmlc/mxnet/pull/6944 Please have a review if it is ok. On Thu, Jul 6, 2017 at 11:17 AM, Suneel Marthi wrote: > it clearly says that's for mini javascripts, images or PDFs - not for > source code. > > "Sometimes the

Re: ZeroMQ licensing in Apache MXNet

MXNet's backend is written in C++, which is not able to use the java interface. On Thu, Jul 6, 2017 at 12:25 PM, Luciano Resende wrote: > Are you guys able to use this (which is what we use in Apache Toree)? > > https://github.com/zeromq/jeromq > > Which has been

Re: ZeroMQ licensing in Apache MXNet

It's optional for MXNet to use ZeroMQ. Even if it is enabled, the source codes of MXNet will not contain any codes from ZeroMQ except for "include" and calling zeromq's APIs. But if we want to ship the binary, it will link against libzeromq.a On Thu, Jul 6, 2017 at 9:21 AM, John D. Ament

Podling Report - May 2017

* Your project name: Apache MXNet * A brief description of your project, which assumes no knowledge of the project or necessarily of its field MXNet is an open-source deep learning framework that allows you to define, train, and deploy deep neural networks on a wide array of devices, from