arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1708.04801cs.LGcs.AIstat.ML
收藏

分布式非平衡负载训练系统的加权并行SGD

Weighted parallel SGD for distributed unbalanced-workload training system

Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所);University of Chinese Academy of Sciences(中国科学院大学)

Cheng Daning, Li Shigang, Zhang Yunquan

摘要:

随机梯度下降(SGD)是机器学习中一种流行的随机优化方法。传统的并行SGD算法,例如,SimuParallel SGD通常要求所有节点具有相同的性能或消耗等量的数据。然而,当并行SGD算法运行在异构计算环境中时,这些要求很难满足,低性能的节点会对最终结果产生负面影响。在本文中,我们提出了一种算法称为加权并行SGD(WP-SGD)。WP-SGD结合来自系统中不同节点的加权模型参数以产生最终输出。WP-SGD利用标准差的减少来补偿集群中节点性能不一致造成的损失,这意味着WP-SGD不要求所有节点消耗等量的数据。我们还分析了理论上的可行性,运行其他两个并行SGD算法结合WP-SGD在异构环境中。实验结果表明,WP-SGD算法在负载不均衡的分布式训练系统上的性能明显优于传统的并行SGD算法。

英文摘要:

Stochastic gradient descent (SGD) is a popular stochastic optimization method in machine learning. Traditional parallel SGD algorithms, e.g., SimuParallel SGD, often require all nodes to have the same performance or to consume equal quantities of data. However, these requirements are difficult to satisfy when the parallel SGD algorithms run in a heterogeneous computing environment; low-performance nodes will exert a negative influence on the final result. In this paper, we propose an algorithm called weighted parallel SGD (WP-SGD). WP-SGD combines weighted model parameters from different nodes in the system to produce the final output. WP-SGD makes use of the reduction in standard deviation to compensate for the loss from the inconsistency in performance of nodes in the cluster, which means that WP-SGD does not require that all nodes consume equal quantities of data. We also analyze the theoretical feasibility of running two other parallel SGD algorithms combined with WP-SGD in a heterogeneous environment. The experimental results show that WP-SGD significantly outperforms the traditional parallel SGD algorithms on distributed training systems with an unbalanced workload.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑