Event-Triggered Gossip for Distributed Learning

Zhiyuan Zhai; Xiaojun Yuan; Wei Ni; Xin Wang; Rui Zhang; Geoffrey Ye Li

eess.SPFeb 22, 2026

Event-Triggered Gossip for Distributed Learning

Authors: Zhiyuan Zhai, Xiaojun Yuan, Wei Ni, Xin Wang, Rui Zhang, Geoffrey Ye Li

Abstract

While distributed learning offers a new learning paradigm for distributed network with no central coordination, it is constrained by communication bottleneck between nodes. We develop a new event-triggered gossip framework for distributed learning to reduce inter-node communication overhead. The framework introduces an adaptive communication control mechanism that enables each node to autonomously decide in a fully decentralized fashion when to exchange model information with its neighbors based on local model deviations. We analyze the ergodic convergence of the proposed framework under noconvex objectives and interpret the convergence guarantees under different triggering conditions. Simulation results show that the proposed framework achieves substantially lower communication overhead than the state-of-the-art distributed learning methods, reducing cumulative point-to-point transmissions by 71.61% with only a marginal performance loss, compared with the conventional full-communication baseline.

Figures & tables

Fig. 1: An example of event-triggered communication model.

Fig. 2: Test accuracy of the averaged model under different triggering thresholds

\epsilon

(MNIST).

Fig. 3: Cumulative point-to-point communication volume versus

\epsilon

(MNIST).

Fig. 4: Test accuracy of the averaged model under different triggering thresholds

\epsilon

(Fashion-MNIST).

Fig. 5: Cumulative point-to-point communication volume versus

\epsilon

(Fashion-MNIST).

Scheme	MNIST		Fashion-MNIST
Scheme	Accuracy	Comm.	Accuracy	Comm.
Full Comm.	0.9219	39900	0.7422	39900
$\epsilon=3\times 10^{-3}$	0.9179	18934	0.7393	17811
$\epsilon=5\times 10^{-3}$	0.9135	12230	0.7360	11328
$\epsilon=7\times 10^{-3}$	0.9087	8893	0.7325	8134
$\epsilon=9\times 10^{-3}$	0.8711	5853	0.7292	6203

TABLE I: Average Accuracy and Cumulative Point-to-Point Communication Volume After 150 Rounds

Fig. 6: Average accuracy under different levels of network sparsity.

Fig. 7: Cumulative point-to-point communication volume under different levels of network sparsity.

Scheme	MNIST		Fashion-MNIST
Scheme	Accuracy	Comm.	Accuracy	Comm.
Full communication	0.9219	39900	0.7422	39900
Event-triggered	0.9135	12230	0.7360	11328
Periodic gossip	0.8454	7980	0.6884	7980
Probabilistic gossip	0.9134	19904	0.7357	19904
Variable working	0.7406	11939	0.6430	11939

TABLE II: Average Accuracy and Cumulative Point-to-Point Communication Volume After 150 Rounds

Fig. 8: Test accuracy versus cumulative point-to-point communication volume for different distributed learning schemes on the MNIST dataset.

Fig. 9: Test accuracy versus cumulative point-to-point communication volume for different distributed learning schemes on the Fashion-MNIST dataset.

Communication-Efficient Distributed Training

Communication-efficient distributed training reduces the data exchanged among devices so models can be trained faster and over limited-bandwidth networks.

See all 227 papers

Stream Learning: Partition-Fair Gossip Learning Without Tokens
Aug 7, 2026Fabien Mathieu, Alexandre Pham, Maria Gradinariu Potop-Butucaru +1Communication-Efficient Distributed Training Decentralized Learning
NPA · Swapcard, Paris, France · Sorbonne Université, CNRS, LIP6, F-75005 Paris, France +2
A Unified Framework for Fair and Personalized Decentralized Learning under Communication Constraints
Aug 27, 2026Krishnendu S. Tharakan, Carlo FischioneDecentralized Learning Communication-Efficient Distributed Training
School of Electrical Engineering and Computer Science, KTH Royal Institute of Technology, Stockholm, Sweden
Communication-Efficient Distributed Learning with Differential Privacy
Apr 2, 2026Xiaoxing Ren, Yuwen Ma, Nicola Bastianello +3Communication-Efficient Distributed Training Differentially Private Federated Learning
School of Civil and Environmental Engineering, Cornell University, Ithaca, New York, United States. · Department of Electronic and Electrical Engineering, University College London, London WC1E 7JE, United Kingdom. · School of Electrical Engineering and Computer Science, and Digital Futures, KTH Royal Institute of Technology, Stockholm, Sweden. +4

Aug 7, 2026J/K move · Enter open · S save

Stream Learning: Partition-Fair Gossip Learning Without Tokens

Fabien Mathieu, Alexandre Pham, Maria Gradinariu Potop-Butucaru, S{é}bastien Tixeuil

NPA · Swapcard, Paris, France · Sorbonne Université, CNRS, LIP6, F-75005 Paris, France+2

Communication-Efficient Distributed Training Decentralized Learning Distributed Training

Open paper AskPDF arXiv

In gossip learning, a network of nodes trains a shared model collaboratively, without a central coordinator, by repeatedly exchanging parts of their local models. The state-of-the-art protocol, Partitioned Token Gossip Learning (PTGL) of Heged{ü}s et al., splits the weight matrix into S fixed partitions and disseminates them using a token-based fairness mechanism coupled with per-neighbor metadata exchange. We revisit partition scheduling by analogy with peer-to-peer live streaming, where model partitions act as video chunks and partition age acts as chunk scarcity. The analogy yields a design space of two-stage selection strategies (partition first, or neighbor first), from which we instantiate ten concrete protocols collectively called Stream Learning. Our main finding is that the simplest of these protocols, which transmits the locally least-trained partition to a uniformly random neighbor (Ri), matches PTGL on fault-free workloads while requiring neither token counters nor metadata exchange. Under an adversarial 30% permanent crash of the best-performing nodes, Ri matches or outperforms PTGL across all complete-graph configurations tested, with the gap reaching 5.53% on HAR and 5.41% on MNIST in the most heterogeneous regime (Dirichlet

β

= 0.1). In our experiments, partition fairness, captured by a single local rule on partition age, accounts for the gap; token-based rate control and utility maximization do not improve over this rule and, under heterogeneity, sit below it.

Abstract

Communication-Efficient Distributed Training

Explore similar work

Stream Learning: Partition-Fair Gossip Learning Without Tokens

A Unified Framework for Fair and Personalized Decentralized Learning under Communication Constraints

Communication-Efficient Distributed Learning with Differential Privacy