mxnet-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Bhavin Thaker <bhavintha...@gmail.com>
Subject Draft MXNet Podling report for Aug-2017
Date Mon, 31 Jul 2017 20:40:12 GMT
Hi All,



Please review the Podling Report for August 2017 so that we can file on
time. Feel free to make updates to this shared Google doc directly.



https://docs.google.com/document/d/1PGhs96klZB6DXhpK9_biPh4-aCm8-bWwFzexnOW_GMA/edit



Given below is the current snapshot of the report – it may change as
updates are made directly to the above shared Google doc.



--snip—

*MXNet*



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

cloud infrastructure to mobile devices. It is highly scalable, allowing for

fast model training, and supports a flexible programming model and multiple

languages. MXNet allows you to mix symbolic and imperative programming

flavors to maximize both efficiency and productivity. MXNet is built on a

dynamic dependency scheduler that automatically parallelizes both symbolic

and imperative operations on the fly. A graph optimization layer on top of

that makes symbolic execution fast and memory efficient. The MXNet library

is portable and lightweight, and it scales to multiple GPUs and multiple

machines.



MXNet has been incubating since 2017-01-23.



*Three most important issues to address in the move towards graduation:*



 1. Migrate code (GitHub) and website to Apache Infra.

 2. Establish a predictable release process consistent with Apache Way.

 3. Grow the community.



*Any issues that the Incubator PMC (IPMC) or ASF Board wish/need to be*

*aware of?*



  None



*How has the community developed since the last report?*



   1. Various Slack channels and dev@ mailing lists are being used actively.
   2. A new blog published on OReilly web-site on 27-July having
   step-by-step instructions to implement a convolutional neural network to
   classify traffic signs with Apache MXNet:

https://www.oreilly.com/ideas/classifying-traffic-signs-with-mxnet-an-introduction-to-computer-vision-with-neural-networks

   1. A new blog post published on 28-July showing users how to exploit the
   unique features of Apache MXNet with a cheat sheet:

https://aws.amazon.com/blogs/ai/exploiting-the-unique-features-of-the-apache-mxnet-deep-learning-framework-with-a-cheat-sheet/



*How has the project developed since the last report?*



   1. The code base was migrated from http://github.com/dmlc/mxnet to
   https://github.com/apache/incubator-mxnet on 28-July, 2017.
   2. From a statistics perspective, 54 authors have pushed 140 commits to
   master, with updates to 358 files including 22K additions and 3K deletions.
   3. Documentation- Architecture guides, How To’s, Tutorials, and APIs
   continue to be improved.
   4. More features (e.g. operators, algorithms) and bug-fixes requested by
   the user community continue to be added.



*How would you assess the podling's maturity?*



  Podling's still getting established in Apache - so maturity == Low.



Please feel free to add your own commentary.

 [X] Initial setup

 [  ] Working towards first release

 [  ] Community building

 [  ] Nearing graduation

 [X] Other: A maintenance release is being planned for August 2017



*Date of last release:*



 A maintenance release MXNet 0.10.0 Post 2 with few bug-fixes was released
on 17-July, 2017.

 https://github.com/apache/incubator-mxnet/releases/tag/0.10.0.post2



*When were the last committers or PPMC members elected?*



 Ly Nguyen added as a committer and PPMC member in June 2017.



*Signed-off-by:*

 [ ](mxnet) Sebastian Schelter

            Comments:

 [ ](mxnet) Suneel Marthi

            Comments:

 [ ](mxnet) Markus Weimer

            Comments:

 [ ](mxnet) Henri Yandell

            Comments:



--snip—



Thanks,

Bhavin Thaker.

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message