arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2006.13878cs.DCcs.LG
收藏

深度学习工作负载的有效弹性缩放

Effective Elastic Scaling of Deep Learning Workloads

IBM Research(IBM研究院);Microsoft(微软)

Vaibhav Saxena, K. R. Jayaram, Saurav Basu, Yogish Sabharwal, Ashish Verma

摘要:

深度学习(DL)在学术界、政府和工业界的使用越来越多,反过来又导致了内部部署和云托管的深度学习平台的普及,其目标是使组织能够有效地利用昂贵的资源,并以公平有效的方式在多个团队之间共享所述资源。在本文中,我们研究了大规模训练平台上深度学习(DL)作业的弹性伸缩,并提出了一种新的DL训练作业的资源分配策略,从而提高了作业运行时间性能,并提高了集群利用率。我们从分析DL工作负载开始,并利用DL作业可以以一系列批量大小运行而不会影响其最终准确性的事实。我们制定了一个优化问题,探讨了一个动态的批量大小分配给个别DL作业的基础上,他们的扩展效率,当运行在多个节点上。我们设计了一个快速的动态规划为基础的优化器来解决这个问题,在实时确定作业,可以按比例增加/减少,并使用此优化器在自动缩放动态改变分配的资源和批量大小的个别DL作业。我们的经验表明,我们的弹性缩放算法可以完成多达$\approx 2 \times$的工作相比,一个强大的基线算法,也规模的GPU数量,但不改变批量大小。我们还证明了我们的算法的平均完成时间是高达$\approx 10 \times$快于基线。

英文摘要:

The increased use of deep learning (DL) in academia, government and industry has, in turn, led to the popularity of on-premise and cloud-hosted deep learning platforms, whose goals are to enable organizations utilize expensive resources effectively, and to share said resources among multiple teams in a fair and effective manner. In this paper, we examine the elastic scaling of Deep Learning (DL) jobs over large-scale training platforms and propose a novel resource allocation strategy for DL training jobs, resulting in improved job run time performance as well as increased cluster utilization. We begin by analyzing DL workloads and exploit the fact that DL jobs can be run with a range of batch sizes without affecting their final accuracy. We formulate an optimization problem that explores a dynamic batch size allocation to individual DL jobs based on their scaling efficiency, when running on multiple nodes. We design a fast dynamic programming based optimizer to solve this problem in real-time to determine jobs that can be scaled up/down, and use this optimizer in an autoscaler to dynamically change the allocated resources and batch sizes of individual DL jobs. We demonstrate empirically that our elastic scaling algorithm can complete up to $\approx 2 \times$ as many jobs as compared to a strong baseline algorithm that also scales the number of GPUs but does not change the batch size. We also demonstrate that the average completion time with our algorithm is up to $\approx 10 \times$ faster than that of the baseline.

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。

↑