Fetching the paper…

High-Performance Distributed ML at Scale through Parameter Server Consistency Models · Around