paper-with-me

홈 › Papers

Optimizing Multi-GPU Parallelization Strategies for Deep Learning Training

2019-07-30 · Saptadeep Pal, Eiman Ebrahimi, Arslan Zulfiqar, Yaosheng Fu, Victor Zhang, Szymon Migacz, David Nellans, Puneet Gupta

Deploying deep learning (DL) models across multiple compute devices to train large and complex models continues to grow in importance because of the demand for faster and more frequent training. Data parallelism (DP) is the most widely used parallelization strategy, but as the number of devices in data parallel training grows, so does the communication overhead between devices. Additionally, a larger aggregate batch size per step leads to statistical efficiency loss, i.e., a larger number of epochs are required to converge to a desired accuracy. These factors affect overall training time and beyond a certain number of devices, the speedup from leveraging DP begins to scale poorly. In addition to DP, each training step can be accelerated by exploiting model parallelism (MP). This work explores hybrid parallelization, where each data parallel worker is comprised of more than one device, across which the model dataflow graph (DFG) is split using MP. We show that at scale, hybrid training will be more effective at minimizing end-to-end training time than exploiting DP alone. We project that for Inception-V3, GNMT, and BigLSTM, the hybrid strategy provides an end-to-end training speedup of at least 26.5%, 8%, and 22% respectively compared to what DP alone can achieve at scale.

📄 PDF Abstract BibTeX arXiv:1907.13257

Code (0)

등록된 구현이 없습니다.

Tasks

Deep LearningGPU

Similar Papers 제목 키워드 기반

Optimal Parallelization Strategies for Active Flow Control in Deep Reinforcement Learning-Based Computational Fluid Dynamics

2024-02-18 · Wang Jia, Hang Xu

Deep Reinforcement Learning (DRL) has emerged as a promising approach for handling highly dynamic and nonlinear Active Flow Control (AFC) problems. However, the computational cost associated with training DRL models pres…

CPUDeep Reinforcement Learning

K-Myriad: Jump-starting reinforcement learning with unsupervised parallel agents

2026-01-26 · Vincenzo De Paola, Mirco Mutti, Riccardo Zamboni, Marcello Restelli arxiv

Parallelization in Reinforcement Learning is typically employed to speed up the training of a single policy, where multiple workers collect experience from an identical sampling distribution. This common design limits th…

Reinforcement LearningContinuous Control

TensorOpt: Exploring the Tradeoffs in Distributed DNN Training with Auto-Parallelism

2020-04-16 · Zhenkun Cai, Kaihao Ma, Xiao Yan, Yidi Wu 외

A good parallelization strategy can significantly improve the efficiency or reduce the cost for the distributed training of deep neural networks (DNNs). Recently, several methods have been proposed to find efficient para…

Parallelization of Monte Carlo Tree Search in Continuous Domains

2020-03-30 · Karl Kurzer, Christoph Hörtnagl, J. Marius Zöllner

Monte Carlo Tree Search (MCTS) has proven to be capable of solving challenging tasks in domains such as Go, chess and Atari. Previous research has developed parallel versions of MCTS, exploiting today's multiprocessing a…

Trajectory Planning

Scalable Synthesis of distributed LLM workloads through Symbolic Tensor Graphs

2025-11-13 · Changhai Man, Joongun Park, Hanjiang Wu, Huan Xu 외 arxiv

Optimizing the performance of large language models (LLMs) on large-scale AI training and inference systems requires a scalable and expressive mechanism to model distributed workload execution. Such modeling is essential…