Model training method and device and cluster system

• Main Authors: ZHANG HENGHUA, LI ZHI, DING RUIQUAN, LUO BAOTONG, HU ZAIBIN, HUANG KAIWEN
• Format: Patent
• Language: Chinese; English
• Published: 23.06.2020
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC COMMUNICATION TECHNIQUE; ELECTRICITY; PHYSICS; TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHICCOMMUNICATION

The embodiment of the invention discloses a model training method and device and a cluster system, and relates to the technical field of artificial intelligence. The specific implementation scheme isas follows: in the aspect of hardware, a control node and at least one computing node are interconnected through a network, and a GPU is introduced into the computing node as a computing resource, sothat the hardware capability of a cluster system is greatly improved, and the efficiency of model training is further improved. In the aspect of software, a surm framework is optimized, and a client,a super management platform and the like are introduced, so that the cluster system is more convenient to use. 本申请实施例公开了一种模型训练方法、装置及集群系统，涉及人工智能技术领域。具体实现方案为：硬件方面，通过将控制节点和至少一个计算节点通过网络互连，在计算节点中引入GPU作为计算资源，从而大幅度提升集群系统的硬件能力，进而提升模型训练的效率。软件方面，通过对slurm框架进行优化，引入客户端、超级管理平台等，使得集群系统用起来更方便。