Apache MXNet (incubating)

Monday March 04, 2019

Apache MXNet 1.4 is available!

Apache MXNet 1.4 is available!

Today the Apache MXNet community is excited to announce the 1.4.0 release of the Apache MXNet deep learning framework. We would like to thank the Apache MXNet community for all their contributions towards this power packed v1.4 release.

This release brings Java bindings for inference, Julia bindings, experimental control flow operators, JVM memory management, graph optimization and quantizations, and many more features and under-the-hood performance enhancements and usability improvements. Check out the full release highlights on MXNet v1.4 release notes.

mxnet-v-1-4-0.png

Java Inference APIs

This release includes new Java Inference APIs which offer easy to use, idiomatic high level APIs for performing predictions in Java with deep learning models trained using MXNet. This simplifies production deployment of Apache MXNet models for enterprise systems that run on Java. More details are available in this blog post which introduces how to use the new Java Inference APIs.

Julia API

The Julia package brings flexible and efficient GPU computing and state-of-the-art deep learning to Julia. Some highlights of the Julia package include efficient tensor computation across multiple devices including multiple CPUs, GPUs and distributed server nodes.

Control Flow Operators (experimental)

With the control flow operators variable dynamic neural network graphs can be turned into optimized static computation graphs. The optimized graph can greatly improve the speed for training and inference for dynamic models. To learn more details about these operators, check out the Control Flow Operators tutorial.

Automated JVM Memory Management

As the name suggests, the memory management feature provides an automated way for managing native memory when using the JVM language bindings of MXNet. Developers now get a seamless memory management system for managing both CPU and GPU memory footprint without any degradation in performance. More details on how to use can be found in this README file.

MXNet Horovod Integration

Apache MXNet now supports distributed training using Horovod framework. Horovod is an open source distributed framework created at Uber. It leverages efficient inter-GPU communication to distribute and aggregate model parameters across multiple workers thus allowing efficient use of network bandwidth and scaling of training of deep learning models. To learn more about MXNet-Horovod integration, check out this blog.

Subgraph API

Subgraph API empowers Apache MXNet to integrate different kinds of backend libraries such as TVM, MKLDNN, TensorRT, Intel nGraph and many more. Enhanced integration with different backend libraries provides MXNet with a significant performance boost, by optimizing the execution of graph by breaking it into smaller components. Check out integrating with external backend libraries to learn more.

MKLDNN : Graph Optimization and Quantization

MKLDNN takes advantage of the MXNet Subgraph API to implement graph optimizations in the form of operator fusion for inference. Fusions such as Convolution + ReLU and Batch Normalization Folding, provide a great boost in the speeds for inference.
Quantization allows the use of reduced precision (INT8), which reduces memory usage and improves inference time without a significant loss in accuracy. On models such as ResNet50, Inception-BN, MobileNet the observed accuracy loss was less than 0.5 %. More details on this project are available on the MXNet Graph Optimization and Quantization page.

Other Maintenance improvements :

  1. The community fixed 81 bugs, improving MXNet's stability and reliability.
  2. The community also addressed over 55 documentation issues, improving the user experience for using MXNet and earning trust of the users.

Getting started with MXNet

Getting started with MXNet is simple, visit the install page to get started. To learn more about MXNet Gluon interface and deep learning, you can follow our 60-minute crash course, and then later complete this comprehensive set of tutorials, which covers everything from an introduction to deep learning to how to implement cutting-edge neural network models. You can check out other related MXNet tutorials, MXNet blog posts and MXNet YouTube channel.

Have fun with MXNet 1.4.0!

Acknowledgements:

We would like to thank everyone from the Apache MXNet community for their contributions to the 1.4.0 release.

Comments:

We would like to thank everyone from the Apache MXNet community for their contributions to the 1.4.0 release.

Posted by film izle on June 29, 2019 at 08:43 PM UTC #

thans for sharing apache site your working is very help full for very user.

Posted by Stephen Warrilow on July 06, 2019 at 02:29 AM UTC #

Oh my God!! Your works are awesome. Nice gift. Nice presentation, good job team :)

Posted by zovirax on July 19, 2019 at 09:34 PM UTC #

Finally! I have been for this release for a long time, thanks! https://eduloh.com/

Posted by EduLoh on July 25, 2019 at 08:53 PM UTC #

Nice gift. Nice presentation, good job team :)

Posted by Fish Tank on September 17, 2019 at 07:23 AM UTC #

Actually, this is one of the best updates, everything works just perfectly, thanks! https://rapidessay.com/custom-thesis.html

Posted by Mary on October 14, 2019 at 08:06 AM UTC #

Woohoo, awesome updates!

Posted by Adkins Masonry on October 17, 2019 at 11:34 PM UTC #

Quick books support is a software and web application company which provide quality softwares and support to the technians who have stuck in their queries.We have a always open to talk customer support center which answers all the queries related to the feild and satisfies every customer by providing quality support and service. Anyone can feel free to talk to us on the- QuickBooks support number- +1 800 786 0194

Posted by Quickbooks on November 20, 2019 at 08:36 AM UTC #

1.4 is one of the best of the latest updates that works stable, no bugs at all Max Kush Developer at https://forexn.com/

Posted by Source on November 29, 2019 at 08:30 AM UTC #

I am certain this is going to help a ton of people. Keep doing awesome. It is profoundly persuading and I delighted in experiencing the whole blog. Regardless of which Australia college you are from, our certified scholars can take into account every one of your needs. In this way, associate with us to put a conclusion to every one of your endeavors. We will be happy to help you. https://assignmentprovideraustralia.com

Posted by Merry Rose on November 29, 2019 at 12:40 PM UTC #

Hey, I am Nora jones. I love to read blogs and articles and do experiments what I learned. I am working at Back link Service as a Technical Expert.

Posted by RussellJackson on December 04, 2019 at 09:24 AM UTC #

Finally, I waited for this update. It remains to update the new plugin for the casino site https://rocketpayz.com/slots-paypal/ . Thanks to you.

Posted by Marvin Kimball on December 05, 2019 at 01:55 PM UTC #

An essay is an interesting task if you have a mood and enough time to write it. It is never the case for me. I have to study and work. So free time is a luxury for me. Luckily, I know a good and https://fastessay.com fast essay writing service. These guys know what they do. They can complete an essay in any theme you need. And you can be sure about the quality. Besides, they will enclose a short summary so that you can get acquainted with the work without problems. Great solution.

Posted by riiupw afrass on December 14, 2019 at 11:31 PM UTC #

https://192-168-1-1.de/ Merci pour l'information utile! Je l'ai aidé vos conseils!

Posted by Sophie Miller on December 19, 2019 at 10:07 AM UTC #

What an update it was, just fantastic, it was really very useful for me and my project https://topcasinoplanet.com/

Posted by George on December 21, 2019 at 10:31 PM UTC #

This is good and amazing job than dear for good info http://productkeyz.com/

Posted by Qais-Saeed on December 25, 2019 at 04:03 PM UTC #

This version is great, I have already tried on my site https://yarkoweb.com/bmw-f30-accessories/

Posted by Yarkoweb on January 15, 2020 at 12:38 PM UTC #

Post a Comment:
  • HTML Syntax: NOT allowed

Calendar

Search

Hot Blogs (today's hits)

Tag Cloud

Categories

Feeds

Links

Navigation