Accomplishments from the first five years of Facebook AI Research (FAIR) | Continue reading
Neural Egg Separation is a new method for identifying distinct images and sounds within noisy environments, a long-standing challenge for ML systems. | Continue reading
FAIR and NYU CAI²R open source tools and data for fastMRI, a joint research project to spur development of AI systems to speed MRI scans by up to 10x. | Continue reading
In one year, predictive test selection has allowed Facebook to double the efficiency of our testing infrastructure and catch 99.9 percent of regressions. | Continue reading
To develop new product features and updates efficiently, we use a trunk-based development model for changes to our codebase. Once an engineer’s code change has been accepted into the main branch (the trunk), we strive to make it visible quickly to every other engineer working on … | Continue reading
Radical changes to floating point math make it as much as 16 percent more efficient than int8/32 math, yet still highly accurate for CNNs. | Continue reading
Facebook is open-sourcing FBGEMM to enable large-scale production servers to run the newest, most powerful deep learning models efficiently. | Continue reading
Facebook researchers developed a new, more accurate zero-shot learning model that uses neural net architectures called GANs to read and analyze text. | Continue reading
Facebook has built a tool called Getafix that automatically finds fixes for bugs and offers them to engineers to approve. This allows engineers to work more effectively, and it promotes better overall code quality. We believe Getafix is the first tool of its kind to be deployed t … | Continue reading
Facebook open-sourced Horizon, an end-to-end applied reinforcement learning platform to optimize products and services used by billions of people. | Continue reading
The XLNI dataset, a collaboration between Facebook and NYU, builds on the MultiNLI corpus, adding 14 languages including low-resource languages. | Continue reading
Katran creates a software-based solution to load balancing with a reengineered forwarding plane that takes advantage of recent innovations in kernel engineering. [...]Read More... | Continue reading
A new suite of open source Linux kernel components and related tools to address critical fleet management issues, inclucing resource control and monitoring. | Continue reading
Facebook open-sources QNNPACK, a high-performance kernel library optimized for mobile AI. QNNPACK speeds up many advanced neural network operations. | Continue reading
Facebook open-sources QNNPACK, a high-performance kernel library optimized for mobile AI. QNNPACK speeds up many advanced neural network operations. | Continue reading
The Symbol-to-Instrument Neural Generator (SING) creates musical notes that sound as realistic as those by similar AI systems — but 2,500 times faster. | Continue reading
MobileLab prevents regressions from shipping and impacting Facebook's mobile performance by detecting tiny changes in performance. | Continue reading
An approach to using word embeddings for NLP, dynamic meta-embeddings outperform similar models. This code is now open source. | Continue reading
At this year's Oculus Connect, we introduced Oculus Quest and shared deep dives into technical challenges and the future of VR. | Continue reading
Facebook’s Safety Check feature has been successful because of its ability to identify those likely to be in an affected area, collect their safety status, and send that information out to friends and family in real time. Safety Check for Workplace incorporates important modifica … | Continue reading
Facebook developed Akkio, a new data placement service (DPS) that operates on trillions of small entities to determine how and when to move information in order to optimize retrieval speed for people across the globe, using the minimum required number of copies. | Continue reading
At Facebook, our services are designed to recover automatically from a major outage, such as the loss of a data center due to a natural disaster. Most of our production services are built in-house and these all run in containers. The third-party services we use in our corporate i … | Continue reading
Facebook Marketplace was introduced in 2016 as a place for people to buy and sell items within their local communities. Today in the U.S., more than one in three people on Facebook use Marketplace,… | Continue reading
Earlier this year, we shared a vision for making AI development faster and more interoperable. Today, during our first-ever PyTorch Developer Conference, we are announcing updates about the growing… | Continue reading
WHAT THE RESEARCH IS: Composable planning is a new way of building AI agents that are better at solving unfamiliar tasks. Traditional training involves agents repeatedly practicing one specific tas… | Continue reading
What the research is: An investigation of omni-supervised learning, a type of semi-supervised learning that uses a combination of data that’s been manually labeled for training purposes (supe… | Continue reading
LogDevice, a distributed data store for sequential data, originally developed and currently deployed within Facebook is now available as an open source project. LogDevice offers durability, high av… | Continue reading
The full article is available at research.fb.com. Facebook relies on a large suite of backend systems to serve billions of people each day. Many of these systems have a large number of internal pa… | Continue reading
Debugging code is drudgery. But SapFix, a new AI hybrid tool created by Facebook engineers, can significantly reduce the amount of time engineers spend on debugging, while also speeding up the proc… | Continue reading
Today, we are announcing the next steps in Facebook’s efforts to build a hardware ecosystem for machine learning (ML) through partner support of the Glow compiler. We’re pleased to anno… | Continue reading
Understanding the text that appears on images is important for improving experiences, such as a more relevant photo search or the incorporation of text into screen readers that make Facebook more a… | Continue reading
WHAT THE RESEARCH IS: The Facebook open switching system (FBOSS) is switch software created specifically to address the needs of a large, rapidly growing network. At ACM SIGCOMM 2018, we shared new… | Continue reading
Automatic language translation is important to Facebook as a way to allow the billions of people who use our services to connect and communicate in their preferred language. To do this well, curren… | Continue reading
Today, we are excited to unveil a new partner ecosystem program called Express Wi-Fi Certified, which allows access point manufacturers to build Wi-Fi hardware that’s compatible with Express … | Continue reading
Larry Zitnick from the Facebook Artificial Intelligence Research (FAIR) group and Daniel Sodickson, M.D., Ph.D, and Michael Recht, M.D., from NYU School of Medicine collaborated on this work. Faceb… | Continue reading
Highly complex services, such as those here at Facebook, have large source code bases in order to deliver a wide range of features and functionality. Even after the machine code for one of these se… | Continue reading
Connectivity is at the heart of Facebook’s work to bring the world closer together. Together with our partners, we’re committed to working on a number of initiatives to build and levera… | Continue reading
High quality internet access helps people communicate, share knowledge, and strengthen their communities and economies — yet billions of people still lack baseline connectivity. To address this pro… | Continue reading
Facebook Open Source | Continue reading
Facebook Open Source | Continue reading
Facebook Open Source | Continue reading
Facebook Open Source | Continue reading
Facebook Open Source | Continue reading
Facebook Open Source | Continue reading
Facebook Open Source | Continue reading
Facebook Open Source | Continue reading