Facebook | Ahmadreza Razian | سید احمدرضا رضیان

Facebook and CUDA Accelerate Deep Learning Research

Published August 11, 2016

Facebook and CUDA Accelerate Deep Learning Research

Last Thursday at the International Conference on Machine Learning (ICML) in New York, Facebook announced a new piece of open source software aimed at streamlining and accelerating deep learning research. The software, named Torchnet, provides developers with a consistent set of widely used deep learning functions and utilities. Torchnet allows developers to write code in a consistent manner speeding development and promoting code re-use both between experiments and across multiple projects.

Torchnet sits atop the popular Torch deep learning framework benefits from GPU acceleration using CUDA and cuDNN.

Torchnet sits atop the popular Torch deep learning framework benefits from GPU acceleration using CUDA and cuDNN. Further, Torchnet has built-in support for asynchronous, parallel data loading and can make full use of multiple GPUs for vastly improved iteration times. This automatic support or multi-GPU training helps Torchnet take full advantage of powerful systems like the NVIDIA DGX-1 with its eight Tesla P100 GPUs.

Facebook and CUDA Accelerate Deep Learning Research

According to the Torchnet research paper, its modular design makes it easy to re-use code in a series of experiments. For instance, running the same experiments on a number of different datasets is accomplished simply by plugging in different dataloaders. And the evaluation criterion can be changed easily by plugging in a different performance meter.

Torchnet adds another powerful tool to data scientists’ toolkit and will help speed the design and training of neural networks, so they can focus on their next great advancement.

Artificial Intelligence Helps the Blind ‘See’ Facebook

Published April 16, 2016

Artificial Intelligence Helps the Blind ‘See’ Facebook

5 April, Facebook introduced a new feature that automatically generates text descriptions of pictures using advanced object recognition technology.

Until now, people using screen readers would only hear the name of the person who shared the photo, followed by the term “photo” when they came upon an image in News Feed. Now they will get a richer description of what’s in a photo. For instance, someone could now hear, “Image may contain three people, smiling, outdoors.”

The Facebook researchers noted that it took nearly ten months to roll the feature out publicly, as they had to train their deep learning models to recognize more than just the people in the images. For instance, since people mostly care about who is in the photo and what they are doing, but sometimes the background of the photo is what makes it interesting or significant.

While that may be intuitive to humans, it is quite challenging to teach a machine to provide as much useful information as possible while acknowledging the social context.

Their neural network models were trained on a million parameters, but they have carefully selected a set of about 100 concepts based on prominence in photos as well as the accuracy of the visual recognition system. They also avoided concepts that had very specific meanings like smiling, jewelry, cars, and boats. Currently, they are ensuring their object detection algorithm on the objects have a minimum precision rate of 0.8.

Facebook Donating 200 GPUs to European Researchers

Published March 19, 2016

Facebook Donating 200 GPUs to European Researchers

The Facebook Artificial Intelligence Research (FAIR) lab announced a new Research Partnership Program to spur advances in Artificial Intelligence and machine learning — Facebook will be giving out 25 servers powered with GPUs, free of charge.

The first recipient to receive 32 GPUs in four GPU servers is Klaus-Robert Müller of TU Berlin. “Dr. Müller will receive four GPU servers that will enable his team to make quicker progress in two research areas: image analysis of breast cancer and chemical modeling of molecules,” FAIR engineering director Serkan Piantino and research lead Florent Perronnin wrote in a blog post.

Facebook will supply server recipients with all of the necessary software to run the servers and will also send some of its own researchers to various institutions for collaboration.

Facebook CEO Mark Zuckerberg made the announcement from Berlin in a live video feed on his Facebook page – he also discussed the latest in AI with Yann LeCun, who leads Facebook AI Research.

ch.

Click on the image above to watch the fireside chat about the announcement.Facebook GPU fireside chat.

“VR is going to need 10 years to become a very mainstream big thing,” Zuckerberg said during the live video feed. “But we’re committed to this. We have the resources to be able to invest and use these investments across the world to bring the research community into this.”

How GPUs are Revolutionizing Machine Learning

Published December 20, 2015

How GPUs are Revolutionizing Machine Learning

NVIDIA announced that Facebook will accelerate its next-generation computing system with the NVIDIA Tesla Accelerated Computing Platform which will enable them to drive a broad range of machine learning applications.

Facebook is the first company to train deep neural networks on the new Tesla M40 GPUs – introduced last month – this will play a large role in their new open source “Big Sur” computing platform, Facebook AI Research’s (FAIR) purpose-built system designed specifically for neural network training.

How GPUs are Revolutionizing Machine Learning—Open Rack V2 compatible 8-GPU server. Big Sur is two times faster than Facebook’s existing system and will enable the company to train twice as many neural networks which in return will help develop more accurate neural network models and new classes of advanced applications.

Training the sophisticated deep neural networks that power applications such as speech translation and autonomous vehicles requires a massive amount of computing performance.

With GPUs accelerating the training times from weeks to hours, it’s not surprising that nearly every leading machine learning researcher and developer is turning to the Tesla Accelerated Computing Platform and the NVIDIA Deep Learning software development kit.

A recent article on WIRED explains how GPUs have proven to be remarkably adept at deep learning and how large web companies like Facebook, Google and Baidu are shifting their computationally intensive applications to GPUs.

The artificial intelligence is on and it’s powered by GPU-accelerated machine learning.

Xbox App on Windows 10: Social Updates Available Today in Beta

Published October 25, 2015

Xbox App on Windows 10: Social Updates Available Today in Beta

At Team Xbox, we want to make it easy to connect with your friends on any device where you play games. This week, we’re excited to start previewing two more ways to expand your social gaming network and share your gaming content with your friends on Xbox Live.

Facebook friend finder: a top requested fan feature, we’re bringing two of your favorite social networks together—Facebook and Xbox Live—and making it easier than ever to find more people to play games and have fun with.

In the Windows 10 Xbox beta app (available in the Windows 10 Store here), look for the Facebook icon in the Suggested Friends area or Settings. Click it to link your Facebook account to your Xbox Live account and start seeing suggestions for friends on Facebook who you can add on Xbox Live. As your friends link their Facebook accounts to their Xbox Live accounts, you’ll see increasing numbers of Facebook friend suggestions, so be sure to spread the word!

Record your voice in your Game DVR clips on your PC: Windows Insider Program and Xbox beta app participants can now plug in a headset and narrate their Windows 10 game clips with their own voice using Game bar and Game DVR hotkeys.

To turn microphone recording on or off for a gaming session, just press Win + Alt + M. Want to turn microphone recording on for all gaming sessions? Go to Game bar > Settings > Audio or Xbox beta app > Settings > Game DVR > Audio. And if you like personalizing your Game DVR hotkeys and want to customize the microphone recording hotkey, do it here: Game bar > Settings > Shortcuts or Xbox beta app > Settings > Game DVR > Keyboard shortcuts.

Share your Windows 10 Store game clips with voice over commentary to your friends on Xbox Live through the Xbox app or any social network or sharing option on your PC.

Also lighting up in the Xbox beta app today are improvements in the way that gamers can find games and redeem codes. Our goal with these changes is to enable quicker and more relevant access to content.

The Store link in the Xbox beta app will now bring you to the Xbox Store—a launching point for browsing and searching for both Windows 10 and Xbox One games. The new interface allows you to see more products at a glance and brings you the latest crop of most popular Windows 10 and Xbox One games, sales and specials, plus your Deals with Gold and Games with Gold games.
In the Xbox One section of the redesigned Xbox Store, gamers will enjoy a persistent search box, reducing clicks between searches. Also, when you type in the name or a description of the Xbox One content that you are searching for, the search box will suggest some top matching search results.
Once you’ve found the content you are looking for, fans will enjoy quicker and more contextual actions. Own the content already? Click the ‘play on console’ button and kick off a game streaming session. Don’t own the content? Just click on the ‘buy’ button to purchase.
Redeeming codes gets easier too! Just like on Xbox One, when codes are sent to you in a message, simply click ‘redeem code’ to redeem. The ‘Redeem a code’ link in the Xbox One section of Xbox Store allows you to enter-25 digit codes right in the app as well.

Check out these new social features arriving in the Windows 10 Xbox beta app today and let us know what you think. Stay tuned for more updates coming to the beta app each month.

See you on Xbox Live!