paper-with-me

Papers

Breadth-First Pipeline Parallelism

2022-11-11 · Joel Lamy-Poirier

We introduce Breadth-First Pipeline Parallelism, a novel training schedule which optimizes the combination of pipeline and data parallelism. Breadth-First Pipeline Parallelism lowers training time, cost and memory usage by combining a high GPU utilization with a small batch size per GPU, and by making use of fully sharded data parallelism. Experimentally, we observed an increase of up to 43% in training throughput for a 52 billion-parameter model using a small batch size per GPU compared to Megatron-LM, which would reduce the training time and cost by the same amount on a large GPU cluster.

📄 PDF Abstract BibTeX arXiv:2211.05953

Code (0)

등록된 구현이 없습니다.

Tasks

GPU

Similar Papers 제목 키워드 기반

Learning Adaptive Parallel Execution for Efficient Code Localization

2026-01-27 · Ke Xu, Siyang Xiao, Ming Liang, Yichen Yu 외 arxiv

Code localization constitutes a key bottleneck in automated software development pipelines. While concurrent tool execution can enhance discovery speed, current agents demonstrate a 34.9% redundant invocation rate, which…

WLB-LLM: Workload-Balanced 4D Parallelism for Large Language Model Training

2025-03-23 · Zheng Wang, Anna Cai, Xinfeng Xie, Zaifeng Pan 외

In this work, we present WLB-LLM, a workLoad-balanced 4D parallelism for large language model training. We first thoroughly analyze the workload imbalance issue in LLM training and identify two primary sources of imbalan…

Language ModelingLanguage ModellingLarge Language Model

AMDP: Asynchronous Multi-Directional Pipeline Parallelism for Large-Scale Models Training

2026-05-28 · Ling Chen, Houming Wu, Wenjie Yu arxiv

Pipeline parallelism is essential for large-scale model training, but existing asynchronous approaches often degrade convergence due to parameter mismatch between forward and backward passes. We propose Asynchronous Mult…

Zero Bubble Pipeline Parallelism

2023-11-30 · Penghui Qi, Xinyi Wan, Guangxing Huang, Min Lin

Pipeline parallelism is one of the key components for large-scale distributed training, yet its efficiency suffers from pipeline bubbles which were deemed inevitable. In this work, we introduce a scheduling strategy that…

Scheduling

HelixPipe: Efficient Distributed Training of Long Sequence Transformers with Attention Parallel Pipeline Parallelism

2025-07-01 · Geng Zhang, Shenggan Cheng, Xuanlei Zhao, Ziming Liu 외 arxiv

As transformer sequence lengths grow, existing pipeline parallelisms incur suboptimal performance due to the quadratic attention computation and the substantial memory overhead. To relieve these challenges, we propose He…