MindTheStep-AsyncPSGD: Adaptive Asynchronous Parallel Stochastic Gradient Descent
Paper in proceeding, 2019

Stochastic Gradient Descent (SGD) is very useful in optimization problems with high-dimensional non-convex target functions, and hence constitutes an important component of several Machine Learning and Data Analytics methods. Recently there have been significant works on understanding the parallelism inherent to SGD, and its convergence properties. Asynchronous, parallel SGD (AsyncPSGD) has received particular attention, due to observed performance benefits. On the other hand, asynchrony implies inherent challenges in understanding the execution of the algorithm and its convergence, stemming from the fact that the contribution of a thread might be based on an old (stale) view of the state. In this work we aim to deepen the understanding of AsyncPSGD in order to increase the statistical efficiency in the presence of stale gradients. We propose new models for capturing the nature of the staleness distribution in a practical setting. Using the proposed models, we derive a staleness-adaptive SGD framework, MindTheStep-AsyncPSGD, for adapting the step size in an online-fashion, which provably reduces the negative impact of asynchrony. Moreover, we provide general convergence time bounds for a wide class of staleness-adaptive step size strategies for convex target functions. We also provide a detailed empirical study, showing how our approach implies faster convergence for deep learning applications.

Author

Karl Bäckström

Chalmers, Computer Science and Engineering (Chalmers), Networks and Systems (Chalmers)

Marina Papatriantafilou

Chalmers, Computer Science and Engineering (Chalmers), Networks and Systems (Chalmers)

Philippas Tsigas

Chalmers, Computer Science and Engineering (Chalmers), Networks and Systems (Chalmers)

Proceedings - 2019 IEEE International Conference on Big Data, Big Data 2019

16-25

2019 IEEE International Conference on Big Data (IEEE Big Data 2019)
Los Angeles, USA,

WASP SAS: Structuring data for continuous processing and ML systems

Wallenberg AI, Autonomous Systems and Software Program, 2018-01-01 -- 2023-01-01.

INDEED

Chalmers, 2016-01-01 -- 2020-12-31.

Future factories in the Cloud (FiC)

Swedish Foundation for Strategic Research (SSF) (GMT14-0032), 2016-01-01 -- 2020-12-31.

Areas of Advance

Information and Communication Technology

Energy

Driving Forces

Sustainable development

Subject Categories

Computer Science

DOI

10.1109/BigData47090.2019.9006054

More information

Latest update

3/21/2023