Hemingway: Modeling Distributed Optimization Algorithms

TL;DR

Hemingway is a proposed ML-based system that models distributed optimization algorithms to automatically select suitable algorithms and cluster sizes, improving performance and convergence in machine learning tasks.

Contribution

The paper introduces Hemingway, a prototype system that models system and convergence characteristics to optimize distributed algorithm selection and cluster sizing.

Findings

Preliminary results demonstrate potential of Hemingway in selecting optimal configurations.

Modeling system and convergence behaviors aids in improving distributed optimization efficiency.

Challenges include accurately capturing system dynamics and convergence patterns.

Abstract

Distributed optimization algorithms are widely used in many industrial machine learning applications. However choosing the appropriate algorithm and cluster size is often difficult for users as the performance and convergence rate of optimization algorithms vary with the size of the cluster. In this paper we make the case for an ML-optimizer that can select the appropriate algorithm and cluster size to use for a given problem. To do this we propose building two models: one that captures the system level characteristics of how computation, communication change as we increase cluster sizes and another that captures how convergence rates change with cluster sizes. We present preliminary results from our prototype implementation called Hemingway and discuss some of the challenges involved in developing such a system.