English
Related papers

Related papers: Clip21: Error Feedback for Gradient Clipping

200 papers

A rich body of prior work has highlighted the existence of communication bottlenecks in synchronous data-parallel training. To alleviate these bottlenecks, a long line of recent work proposes gradient and model compression methods. In this…

Distributed, Parallel, and Cluster Computing · Computer Science 2021-07-01 Saurabh Agarwal , Hongyi Wang , Shivaram Venkataraman , Dimitris Papailiopoulos

Distributed model training suffers from communication bottlenecks due to frequent model updates transmitted across compute nodes. To alleviate these bottlenecks, practitioners use gradient compression techniques like sparsification,…

Machine Learning · Computer Science 2020-11-02 Saurabh Agarwal , Hongyi Wang , Kangwook Lee , Shivaram Venkataraman , Dimitris Papailiopoulos

In distributed or federated optimization and learning, communication between the different computing units is often the bottleneck and gradient compression is widely used to reduce the number of bits sent within each communication round of…

Machine Learning · Computer Science 2023-03-07 Laurent Condat , Kai Yi , Peter Richtárik

Differentially Private Stochastic Gradient Descent (DPSGD) is widely utilized to preserve training data privacy in deep learning, which first clips the gradients to a predefined norm and then injects calibrated noise into the training…

Machine Learning · Computer Science 2024-05-29 Haichao Sha , Yang Cao , Yong Liu , Yuncheng Wu , Ruixuan Liu , Hong Chen

Sign-based algorithms (e.g. signSGD) have been proposed as a biased gradient compression technique to alleviate the communication bottleneck in training large neural networks across multiple workers. We show simple convex counter-examples…

Machine Learning · Computer Science 2019-05-30 Sai Praneeth Karimireddy , Quentin Rebjock , Sebastian U. Stich , Martin Jaggi

U-Clip is a simple amendment to gradient clipping that can be applied to any iterative gradient optimization algorithm. Like regular clipping, U-Clip involves using gradients that are clipped to a prescribed size (e.g. with component wise…

Machine Learning · Computer Science 2023-02-07 Bryn Elesedy , Marcus Hutter

This paper addresses two fundamental challenges in distributed online convex optimization: communication efficiency and optimization under limited feedback. We propose a unified framework named Online Compressed Gradient Tracking (OCGT),…

Optimization and Control · Mathematics 2025-12-08 Longkang Zhu , Xinli Shi , Xiangping Xu , Jinde Cao , Xiangyong Chen

Recent advances have substantially improved the accuracy, memory cost, and training speed of differentially private (DP) deep learning, especially on large vision and language models with millions to billions of parameters. In this work, we…

Machine Learning · Computer Science 2023-10-31 Zhiqi Bu , Ruixuan Liu , Yu-Xiang Wang , Sheng Zha , George Karypis

We propose and study a new class of gradient communication mechanisms for communication-efficient training -- three point compressors (3PC) -- as well as efficient distributed nonconvex optimization algorithms that can take advantage of…

Machine Learning · Computer Science 2022-02-03 Peter Richtárik , Igor Sokolov , Ilyas Fatkhullin , Elnur Gasanov , Zhize Li , Eduard Gorbunov

Differentially Private Stochastic Gradient Descent with Gradient Clipping (DPSGD-GC) is a powerful tool for training deep learning models using sensitive data, providing both a solid theoretical privacy guarantee and high efficiency.…

Machine Learning · Computer Science 2024-04-18 Xinwei Zhang , Zhiqi Bu , Zhiwei Steven Wu , Mingyi Hong

This paper addresses a distributed convex optimization problem with a class of coupled constraints, which arise in a multi-agent system composed of multiple communities modeled by cliques. First, we propose a fully distributed…

Optimization and Control · Mathematics 2022-11-21 Yuto Watanabe , Kazunori Sakurama

Communication overhead is the key challenge for distributed training. Gradient compression is a widely used approach to reduce communication traffic. When combining with parallel communication mechanism method like pipeline, gradient…

Machine Learning · Computer Science 2021-09-08 Enda Yu , Dezun Dong , Yemao Xu , Shuo Ouyang , Xiangke Liao

Large neural networks require enormous computational clusters of machines. Model-parallel training, when the model architecture is partitioned sequentially between workers, is a popular approach for training modern models. Information…

Machine Learning · Computer Science 2024-03-27 Mikhail Rudakov , Aleksandr Beznosikov , Yaroslav Kholodov , Alexander Gasnikov

We analyze (stochastic) gradient descent (SGD) with delayed updates on smooth quasi-convex and non-convex functions and derive concise, non-asymptotic, convergence rates. We show that the rate of convergence in all cases consists of two…

Machine Learning · Computer Science 2021-06-17 Sebastian U. Stich , Sai Praneeth Karimireddy

In this work, we study the convergence \emph{in high probability} of clipped gradient methods when the noise distribution has heavy tails, ie., with bounded $p$th moments, for some $1<p\le2$. Prior works in this setting follow the same…

Optimization and Control · Mathematics 2023-04-05 Ta Duy Nguyen , Alina Ene , Huy L. Nguyen

Gradient clipping is a commonly used technique to stabilize the training process of neural networks. A growing body of studies has shown that gradient clipping is a promising technique for dealing with the heavy-tailed behavior that emerged…

Machine Learning · Computer Science 2023-07-26 Shaojie Li , Yong Liu

Due to the explosion in the size of the training datasets, distributed learning has received growing interest in recent years. One of the major bottlenecks is the large communication cost between the central server and the local workers.…

Machine Learning · Computer Science 2022-02-25 Yujia Wang , Lu Lin , Jinghui Chen

Differentially private (DP) training preserves the data privacy usually at the cost of slower convergence (and thus lower accuracy), as well as more severe mis-calibration than its non-private counterpart. To analyze the convergence of DP…

Machine Learning · Computer Science 2023-06-21 Zhiqi Bu , Hua Wang , Zongyu Dai , Qi Long

We establish the O($\frac{1}{k}$) convergence rate for distributed stochastic gradient methods that operate over strongly convex costs and random networks. The considered class of methods is standard each node performs a weighted average of…

Optimization and Control · Mathematics 2018-03-22 Dusan Jakovetic , Dragana Bajovic , Anit Kumar Sahu , Soummya Kar

Parallel stochastic gradient methods are gaining prominence in solving large-scale machine learning problems that involve data distributed across multiple nodes. However, obtaining unbiased stochastic gradients, which have been the focus of…

Machine Learning · Computer Science 2025-01-14 Ali Beikmohammadi , Sarit Khirirat , Sindri Magnússon