PowerGossip: Practical Low-Rank Communication Compression in Decentralized Deep Learning

arXiv2008.01425MaRDI QIDQ6346456

Author name not available (Why is that?)

Publication date: 4 August 2020

Abstract: Lossy gradient compression has become a practical tool to overcome the communication bottleneck in centrally coordinated distributed training of machine learning models. However, algorithms for decentralized training with compressed communication over arbitrary connected networks have been more complicated, requiring additional memory and hyperparameters. We introduce a simple algorithm that directly compresses the model differences between neighboring workers using low-rank linear compressors applied on model differences. Inspired by the PowerSGD algorithm for centralized deep learning, this algorithm uses power iteration steps to maximize the information transferred per bit. We prove that our method requires no additional hyperparameters, converges faster than prior methods, and is asymptotically independent of both the network and the compression. Out of the box, these compressors perform on par with state-of-the-art tuned compression algorithms in a series of deep learning benchmarks.

Has companion code repository: https://github.com/epfml/powergossip

This page was built for publication: PowerGossip: Practical Low-Rank Communication Compression in Decentralized Deep Learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6346456)