English
Related papers

Related papers: DeMo: Decoupled Momentum Optimization

200 papers

The training of large models demands substantial computational resources, typically available only in data centers with high-bandwidth interconnects. However, reducing the reliance on high-bandwidth interconnects between nodes enables the…

Machine Learning · Computer Science 2025-10-07 Sasho Nedelkoski , Alexander Acker , Odej Kao , Soeren Becker , Dominik Scheinert

Training large neural network models requires extensive computational resources, often distributed across several nodes and accelerators. Recent findings suggest that it may be sufficient to only exchange the fast moving components of the…

Machine Learning · Computer Science 2025-11-18 Mogens Henrik From , Jacob Nielsen , Lukas Galke Poech , Peter Schneider-Kamp

Distributed optimization is essential for training large models on large datasets. Multiple approaches have been proposed to reduce the communication overhead in distributed training, such as synchronizing only after performing multiple…

Machine Learning · Computer Science 2020-02-21 Jianyu Wang , Vinayak Tantia , Nicolas Ballas , Michael Rabbat

Training large models with distributed data parallelism (DDP) requires frequent communication of gradients across workers, which can saturate bandwidth. Infrequent communication strategies (e.g., Local SGD) reduce this overhead but, when…

Scaling foundation model training with Distributed Data Parallel (DDP) methods is bandwidth-limited. Existing infrequent communication methods like Local SGD were designed to synchronize only model parameters and cannot be trivially applied…

Modern large-scale language model pre-training relies heavily on the single program multiple data (SPMD) paradigm, which requires tight coupling across accelerators. Due to this coupling, transient slowdowns, hardware failures, and…

This paper introduces Decoupled Supervised Learning with Information Regularization (DeInfoReg), a novel approach that transforms a long gradient flow into multiple shorter ones, thereby mitigating the vanishing gradient problem.…

Machine Learning · Computer Science 2025-07-16 Zih-Hao Huang , You-Teng Lin , Hung-Hsuan Chen

We introduce a memory- and compute-efficient method for low-communication distributed training. Existing methods reduce communication by performing multiple local updates between infrequent global synchronizations. We demonstrate that their…

Machine Learning · Computer Science 2025-09-29 Anastasiia Filippova , Angelos Katharopoulos , David Grangier , Ronan Collobert

In this paper, we design two compressed decentralized algorithms for solving nonconvex stochastic optimization under two different scenarios. Both algorithms adopt a momentum technique to achieve fast convergence and a message-compression…

Machine Learning · Computer Science 2025-08-08 Wei Liu , Anweshit Panda , Ujwal Pandey , Christopher Brissette , Yikang Shen , George M. Slota , Naigang Wang , Jie Chen , Yangyang Xu

To efficiently train large-scale models, low-bit gradient communication compresses full-precision gradients on local GPU nodes into low-precision ones for higher gradient synchronization efficiency among GPU nodes. However, it often…

Machine Learning · Computer Science 2024-12-02 Xingyu Xie , Zhijie Lin , Kim-Chuan Toh , Pan Zhou

While Mixture-of-Experts (MoE) scales model capacity without proportionally increasing computation, its massive total parameter footprint creates significant storage and memory-access bottlenecks, which hinder efficient end-side deployment…

Machine Learning · Computer Science 2026-05-21 Chenyang Song , Weilin Zhao , Xu Han , Chaojun Xiao , Yingfa Chen , Zhiyuan Liu

Momentum-Aided Prompt Optimization (MAPO) enhances the efficiency and efficacy of prompt optimization for Large Language Models (LLMs). Building on ProTeGi, MAPO uses positive natural language "gradients" and a momentum-based extension to…

Computation and Language · Computer Science 2025-06-30 Anthony Cui , Pranav Nandyalam , Andrew Rufail , Ethan Cheung , Aiden Lei , Kevin Zhu , Sean O'Brien

Communication scheduling aims to reduce communication bottlenecks in data parallel training (DP) by maximizing the overlap between computation and communication. However, existing schemes fall short due to three main issues: (1) hard data…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-03-24 Lin Meng , Yuzhong Sun

Distributed training of deep neural networks has received significant research interest, and its major approaches include implementations on multiple GPUs and clusters. Parallelization can dramatically improve the efficiency of training…

Distributed, Parallel, and Cluster Computing · Computer Science 2017-11-29 Jaehee Jang , Byungook Na , Sungroh Yoon

Scaling models has led to significant advancements in deep learning, but training these models in decentralized settings remains challenging due to communication bottlenecks. While existing compression techniques are effective in…

Machine Learning · Computer Science 2025-06-03 Sameera Ramasinghe , Thalaiyasingam Ajanthan , Gil Avraham , Yan Zuo , Alexander Long

Low-rank gradient-based optimization methods have significantly improved memory efficiency during the training of large language models (LLMs), enabling operations within constrained hardware without sacrificing performance. However, these…

Machine Learning · Computer Science 2025-10-28 Yehonathan Refael , Guy Smorodinsky , Tom Tirer , Ofir Lindenbaum

Communication efficiency has garnered significant attention as it is considered the main bottleneck for large-scale decentralized Machine Learning applications in distributed and federated settings. In this regime, clients are restricted to…

Machine Learning · Computer Science 2024-11-26 Rustem Islamov , Yuan Gao , Sebastian U. Stich

Decentralized optimization methods enable on-device training of machine learning models without a central coordinator. In many scenarios communication between devices is energy demanding and time consuming and forms the bottleneck of the…

Optimization and Control · Mathematics 2020-11-04 Dmitry Kovalev , Anastasia Koloskova , Martin Jaggi , Peter Richtarik , Sebastian U. Stich

Decentralized stochastic optimization has emerged as a fundamental paradigm for large-scale machine learning. However, practical implementations often rely on biased gradient estimators arising from communication compression or inexact…

Optimization and Control · Mathematics 2026-04-10 Qing Xu , Yiwei Liao , Wenqi Fan , Xingxing You , Songyi Dian

Massive multiuser (MU) multiple-input multiple-output (MIMO) promises significant improvements in spectral efficiency compared to small-scale MIMO. Typical massive MU-MIMO base-station (BS) designs rely on centralized linear data detectors…

Distributed, Parallel, and Cluster Computing · Computer Science 2019-02-26 Kaipeng Li , Oscar Castaneda , Charles Jeon , Joseph R. Cavallaro , Christoph Studer
‹ Prev 1 2 3 10 Next ›