Loading Now

Summary of Cohort Squeeze: Beyond a Single Communication Round Per Cohort in Cross-device Federated Learning, by Kai Yi et al.


Cohort Squeeze: Beyond a Single Communication Round per Cohort in Cross-Device Federated Learning

by Kai Yi, Timur Kharisov, Igor Sokolov, Peter Richtárik

First submitted to arxiv on: 3 Jun 2024

Categories

  • Main: Machine Learning (cs.LG)
  • Secondary: None

     Abstract of paper      PDF of paper


GrooveSquid.com Paper Summaries

GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!

Summary difficulty Written by Summary
High Paper authors High Difficulty Summary
Read the original abstract here
Medium GrooveSquid.com (original content) Medium Difficulty Summary
This paper investigates whether it’s possible to improve federated learning methods that operate by sending model parameters to clients, having them perform local training, and then aggregating the results. The authors challenge this design primitive and find a way to “squeeze more juice” out of each cohort, reducing the total communication cost needed to train a federated learning model in the cross-device setting. This approach uses a novel variant of the stochastic proximal point method (SPPM-AS) that supports different client sampling procedures, leading to further gains compared to classical approaches.
Low GrooveSquid.com (original content) Low Difficulty Summary
Federated learning is a way for many devices to work together without sharing their personal data. Right now, most methods send model parameters to some devices, have them do some local training, and then combine the results. The authors thought, “What if we could make each device do more before sending back its results?” They found that by doing things differently, they could reduce the amount of communication needed to train a model by up to 74%! This is important because it means devices can work together more efficiently.

Keywords

» Artificial intelligence  » Federated learning