Breakthrough in 'distributed deep learning'

ScienceDaily
Dec 9, 2019, 4:13 pm170 pts
Computer scientists, using a divide-and-conquer approach that leverages the power of compressed sensing, have shown they can train the equivalent of a 100 billion-parameter distributed deep learning network on a single machine in less than 35 hours for product search and similar extreme classification problems.

Read Article Share
Share Article
- email
- x.com
- facebook
- pocket
- reddit
- tumblr
- linkedin
- pinterest

Trending Today on EcoTopical

A Rising Tide Lifted Floating Wind Turbines. Now Comes the Real Test. 85

The Most Annoying Neighbor in This San Antonio Subdivision Is a Leaking Oil Well 78

Electric Vehicle Chargers Or Flying Cars? Tesla And The State Of Florida Can’t Decide 18

About EcoTopical

Welcome to EcoTopical Your daily eco-friendly green news aggregator.

Leaf through planet Earths environmental headlines in one convenient place. Read, share and discover the latest on ecology, science and green living from the web's most popular sites.