English
Related papers

Related papers: Accelerating Decentralized Optimization via Overla…

200 papers

Decentralization is a promising method of scaling up parallel machine learning systems. In this paper, we provide a tight lower bound on the iteration complexity for such methods in a stochastic non-convex setting. Our lower bound reveals a…

Machine Learning · Computer Science 2022-01-31 Yucheng Lu , Christopher De Sa

When scaling distributed training, the communication overhead is often the bottleneck. In this paper, we propose a novel SGD variant with reduced communication and adaptive learning rates. We prove the convergence of the proposed algorithm…

Machine Learning · Computer Science 2020-12-08 Cong Xie , Oluwasanmi Koyejo , Indranil Gupta , Haibin Lin

With increasing data and model complexities, the time required to train neural networks has become prohibitively large. To address the exponential rise in training time, users are turning to data parallel neural networks (DPNN) to utilize…

Machine Learning · Computer Science 2022-02-09 Daniel Coquelin , Charlotte Debus , Markus Götz , Fabrice von der Lehr , James Kahn , Martin Siggel , Achim Streit

Decentralized federated learning (DFL) captures FL settings where both (i) model updates and (ii) model aggregations are exclusively carried out by the clients without a central server. Existing DFL works have mostly focused on settings…

Machine Learning · Computer Science 2025-03-07 Shahryar Zehtabi , Dong-Jun Han , Rohit Parasnis , Seyyedali Hosseinalipour , Christopher G. Brinton

Distributed training of deep learning models on large-scale training data is typically conducted with asynchronous stochastic optimization to maximize the rate of updates, at the cost of additional noise introduced from asynchrony. In…

Distributed, Parallel, and Cluster Computing · Computer Science 2017-03-21 Xinghao Pan , Jianmin Chen , Rajat Monga , Samy Bengio , Rafal Jozefowicz

Distributed training of deep learning models on large-scale training data is typically conducted with asynchronous stochastic optimization to maximize the rate of updates, at the cost of additional noise introduced from asynchrony. In…

Machine Learning · Computer Science 2017-03-22 Jianmin Chen , Xinghao Pan , Rajat Monga , Samy Bengio , Rafal Jozefowicz

In decentralized optimization over networks, synchronizing the updates of all nodes incurs significant communication overhead. For this reason, much of the recent literature has focused on the analysis and design of asynchronous…

Optimization and Control · Mathematics 2025-04-29 Marina Costantini , Nikolaos Liakopoulos , Panayotis Mertikopoulos , Thrasyvoulos Spyropoulos

We propose a novel variant of SGD customized for training network architectures that support anytime behavior: such networks produce a series of increasingly accurate outputs over time. Efficient architectural designs for these networks…

Machine Learning · Computer Science 2020-08-18 Chengcheng Wan , Henry Hoffmann , Shan Lu , Michael Maire

We study diffusion and consensus based optimization of a sum of unknown convex objective functions over distributed networks. The only access to these functions is through stochastic gradient oracles, each of which is only available at a…

Numerical Analysis · Computer Science 2015-09-01 N. Denizcan Vanli , Muhammed O. Sayin , Suleyman S. Kozat

We analyze the convergence of gradient-based optimization algorithms that base their updates on delayed stochastic gradient information. The main application of our results is to the development of gradient-based distributed optimization…

Optimization and Control · Mathematics 2011-05-02 Alekh Agarwal , John C. Duchi

This paper proposes a communication strategy for decentralized learning on wireless systems. Our discussion is based on the decentralized parallel stochastic gradient descent (D-PSGD), which is one of the state-of-the-art algorithms for…

Networking and Internet Architecture · Computer Science 2020-02-26 Koya Sato , Yasuyuki Satoh , Daisuke Sugimura

Distributed gradient descent (DGD) is an efficient way of implementing gradient descent (GD), especially for large data sets, by dividing the computation tasks into smaller subtasks and assigning to different computing servers (CSs) to be…

Information Theory · Computer Science 2018-11-29 Emre Ozfatura , Deniz Gunduz , Sennur Ulukus

Neural networks are usually trained by some form of stochastic gradient descent (SGD)). A number of strategies are in common use intended to improve SGD optimization, such as learning rate schedules, momentum, and batching. These are…

Neural and Evolutionary Computing · Computer Science 2015-08-13 Thomas M. Breuel

In distributed and federated learning algorithms, communication overhead is often reduced by performing multiple local updates between communication rounds. However, due to data heterogeneity across nodes and the local gradient noise within…

Machine Learning · Computer Science 2025-12-02 Yan Huang , Jinming Xu , Jiming Chen , Karl Henrik Johansson

We consider the problem of decentralized optimization in networks with communication delays. To accommodate delays, we need decentralized optimization algorithms that work on directed graphs. Existing approaches require nodes to know their…

Optimization and Control · Mathematics 2024-12-31 Tomas Ortega , Hamid Jafarkhani

This paper addresses decentralized stochastic gradient descent (D-SGD) over resource-constrained networks by introducing node-based and link-based scheduling strategies to enhance communication efficiency. In each iteration of the D-SGD…

Information Theory · Computer Science 2025-09-16 Jaiprakash Nagar , Zheng Chen , Marios Kountouris , Photios A. Stavrou

Understanding the bottlenecks in implementing stochastic gradient descent (SGD)-based distributed support vector machines (SVM) algorithm is important in training larger data sets. The communication time to do the model synchronization…

Distributed, Parallel, and Cluster Computing · Computer Science 2019-05-06 Vibhatha Abeykoon , Geoffrey Fox , Minje Kim

Artificial intelligence has advanced rapidly through large neural networks trained on massive datasets using thousands of GPUs or TPUs. Such training can occupy entire data centers for weeks and requires enormous computational and energy…

Optimization and Control · Mathematics 2026-01-07 Artavazd Maranjyan

Distributed training methods are crucial for large language models (LLMs). However, existing distributed training methods often suffer from communication bottlenecks, stragglers, and limited elasticity, particularly in heterogeneous or…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-02-18 Jialiang Cheng , Ning Gao , Yun Yue , Zhiling Ye , Jiadi Jiang , Jian Sha

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…