Petuum: A New Platform for Distributed Machine Learning on Big Data
This work provides a universal platform for scalable distributed ML, addressing a broad problem for researchers and practitioners dealing with big data and models, though it builds incrementally on existing parallelization strategies.
The authors tackled the challenge of efficiently applying diverse advanced ML programs to industrial-scale problems with big models and big data by proposing Petuum, a general-purpose distributed platform that addresses data- and model-parallel challenges through error-tolerant, iterative-convergent designs, resulting in faster runtimes and support for larger model sizes on modest clusters.
What is a systematic way to efficiently apply a wide spectrum of advanced ML programs to industrial scale problems, using Big Models (up to 100s of billions of parameters) on Big Data (up to terabytes or petabytes)? Modern parallelization strategies employ fine-grained operations and scheduling beyond the classic bulk-synchronous processing paradigm popularized by MapReduce, or even specialized graph-based execution that relies on graph representations of ML programs. The variety of approaches tends to pull systems and algorithms design in different directions, and it remains difficult to find a universal platform applicable to a wide range of ML programs at scale. We propose a general-purpose framework that systematically addresses data- and model-parallel challenges in large-scale ML, by observing that many ML programs are fundamentally optimization-centric and admit error-tolerant, iterative-convergent algorithmic solutions. This presents unique opportunities for an integrative system design, such as bounded-error network synchronization and dynamic scheduling based on ML program structure. We demonstrate the efficacy of these system designs versus well-known implementations of modern ML algorithms, allowing ML programs to run in much less time and at considerably larger model sizes, even on modestly-sized compute clusters.