Yujie Wang, Shenhan Zhu, Fangcheng Fu, Xupeng Miao, Jie Zhang, Juan Zhu, Fan Hong, Yong Li, Bin Cui
{"title":"通过数据异质性感知模型管理实现高效的多任务大型模型训练","authors":"Yujie Wang, Shenhan Zhu, Fangcheng Fu, Xupeng Miao, Jie Zhang, Juan Zhu, Fan Hong, Yong Li, Bin Cui","doi":"arxiv-2409.03365","DOIUrl":null,"url":null,"abstract":"Recent foundation models are capable of handling multiple machine learning\n(ML) tasks and multiple data modalities with the unified base model structure\nand several specialized model components. However, the development of such\nmulti-task (MT) multi-modal (MM) models poses significant model management\nchallenges to existing training systems. Due to the sophisticated model\narchitecture and the heterogeneous workloads of different ML tasks and data\nmodalities, training these models usually requires massive GPU resources and\nsuffers from sub-optimal system efficiency. In this paper, we investigate how to achieve high-performance training of\nlarge-scale MT MM models through data heterogeneity-aware model management\noptimization. The key idea is to decompose the model execution into stages and\naddress the joint optimization problem sequentially, including both\nheterogeneity-aware workload parallelization and dependency-driven execution\nscheduling. Based on this, we build a prototype system and evaluate it on\nvarious large MT MM models. Experiments demonstrate the superior performance\nand efficiency of our system, with speedup ratio up to 71% compared to\nstate-of-the-art training systems.","PeriodicalId":501422,"journal":{"name":"arXiv - CS - Distributed, Parallel, and Cluster Computing","volume":"2 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Efficient Multi-Task Large Model Training via Data Heterogeneity-aware Model Management\",\"authors\":\"Yujie Wang, Shenhan Zhu, Fangcheng Fu, Xupeng Miao, Jie Zhang, Juan Zhu, Fan Hong, Yong Li, Bin Cui\",\"doi\":\"arxiv-2409.03365\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Recent foundation models are capable of handling multiple machine learning\\n(ML) tasks and multiple data modalities with the unified base model structure\\nand several specialized model components. However, the development of such\\nmulti-task (MT) multi-modal (MM) models poses significant model management\\nchallenges to existing training systems. Due to the sophisticated model\\narchitecture and the heterogeneous workloads of different ML tasks and data\\nmodalities, training these models usually requires massive GPU resources and\\nsuffers from sub-optimal system efficiency. In this paper, we investigate how to achieve high-performance training of\\nlarge-scale MT MM models through data heterogeneity-aware model management\\noptimization. The key idea is to decompose the model execution into stages and\\naddress the joint optimization problem sequentially, including both\\nheterogeneity-aware workload parallelization and dependency-driven execution\\nscheduling. Based on this, we build a prototype system and evaluate it on\\nvarious large MT MM models. Experiments demonstrate the superior performance\\nand efficiency of our system, with speedup ratio up to 71% compared to\\nstate-of-the-art training systems.\",\"PeriodicalId\":501422,\"journal\":{\"name\":\"arXiv - CS - Distributed, Parallel, and Cluster Computing\",\"volume\":\"2 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-05\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Distributed, Parallel, and Cluster Computing\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.03365\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Distributed, Parallel, and Cluster Computing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.03365","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
最近的基础模型能够通过统一的基础模型结构和多个专用模型组件处理多种机器学习(ML)任务和多种数据模式。然而,这种多任务(MT)多模态(MM)模型的开发给现有的训练系统带来了巨大的模型管理挑战。由于复杂的模型架构以及不同 ML 任务和数据模态的异构工作负载,训练这些模型通常需要大量 GPU 资源,而且系统效率未达到最佳。在本文中,我们研究了如何通过数据异构感知模型管理优化来实现大规模 MT MM 模型的高性能训练。其关键思路是将模型执行分解为若干阶段,并按顺序解决联合优化问题,包括异构感知工作负载并行化和依赖驱动的执行调度。在此基础上,我们构建了一个原型系统,并在各种大型 MT MM 模型上对其进行了评估。实验证明了我们系统的卓越性能和效率,与最先进的训练系统相比,提速比高达 71%。
Efficient Multi-Task Large Model Training via Data Heterogeneity-aware Model Management
Recent foundation models are capable of handling multiple machine learning
(ML) tasks and multiple data modalities with the unified base model structure
and several specialized model components. However, the development of such
multi-task (MT) multi-modal (MM) models poses significant model management
challenges to existing training systems. Due to the sophisticated model
architecture and the heterogeneous workloads of different ML tasks and data
modalities, training these models usually requires massive GPU resources and
suffers from sub-optimal system efficiency. In this paper, we investigate how to achieve high-performance training of
large-scale MT MM models through data heterogeneity-aware model management
optimization. The key idea is to decompose the model execution into stages and
address the joint optimization problem sequentially, including both
heterogeneity-aware workload parallelization and dependency-driven execution
scheduling. Based on this, we build a prototype system and evaluate it on
various large MT MM models. Experiments demonstrate the superior performance
and efficiency of our system, with speedup ratio up to 71% compared to
state-of-the-art training systems.