FuncPipe: A Pipelined Serverless Framework for Fast and Cost-Efficient Training of Deep Learning Models
Open Access
- 8 December 2022
- journal article
- research article
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Measurement and Analysis of Computing Systems
- Vol. 6 (3) , 1-30
Abstract
Training deep learning (DL) models in the cloud has become a norm. With the emergence of serverless computing and its benefits of true pay-as-you-go pricing and scalability, systems researchers have recently started to provide support for serverless-based training. However, the ability to train DL models on serverless platforms is hindered by the resource limitations of today's serverless infrastructure and DL models' explosive requirement for memory and bandwidth. This paper describes FuncPipe, a novel pipelined training framework specifically designed for serverless platforms that enable fast and low-cost training of DL models. FuncPipe is designed with the key insight that model partitioning can be leveraged to bridge both memory and bandwidth gaps between the capacity of serverless functions and the requirement of DL training. Conceptually simple, we have to answer several design questions, including how to partition the model, configure each serverless function, and exploit each function's uplink/downlink bandwidth. In particular, we tailor a micro-batch scheduling policy for the serverless environment, which serves as the basis for the subsequent optimization. Our Mixed-Integer Quadratic Programming formulation automatically and simultaneously configures serverless resources and partitions models to fit within the resource constraints. Lastly, we improve the bandwidth efficiency of storage-based synchronization with a novel pipelined scatter-reduce algorithm. We implement FuncPipe on two popular cloud serverless platforms and show that it achieves 7%-77% cost savings and 1.3X-2.2X speedup compared to state-of-the-art serverless-based frameworks.Keywords
All Related Versions
Funding Information
- VMware
- Natural Science Foundation of China (No. 42050105, 62072302, 62020106005, 62061146002, 61960206002)
- US National Science Foundation (NGSDI-2105564)
- Program of Shanghai Academic/Technology Research Leader (No. 18XD1401800)
This publication has 28 references indexed in Scilit:
- Narrowing the Gap Between Serverless and its State with Storage Functions2019
- PipeDream2019
- Horizontal or Vertical?2019
- Distributed Machine Learning with a Serverless Architecture2019
- Exploring Serverless Computing for Neural Network Training2018
- Serverless Computing: Design, Implementation, and Performance2017
- Bandwidth optimal all-reduce algorithms for clusters of workstationsJournal of Parallel and Distributed Computing, 2009
- Bandwidth Efficient All-reduce Operation on Tree Topologies2007
- Pareto Multi Objective Optimization2006
- Optimization of Collective Communication Operations in MPICHThe International Journal of High Performance Computing Applications, 2005