Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Federated Learning Playground

Bryan Guanrong Shan, Alysa Ziying Tan, Han Yu

TL;DR

Federated Learning Playground is an interactive browser-based platform inspired by and extends TensorFlow Playground that teaches core Federated Learning concepts, lowering the entry barrier for newcomers to distributed AI while also offering a sandbox for rapidly prototyping and comparing FL methods.

Abstract

We present Federated Learning Playground, an interactive browser-based platform inspired by and extends TensorFlow Playground that teaches core Federated Learning (FL) concepts. Users can experiment with heterogeneous client data distributions, model hyperparameters, and aggregation algorithms directly in the browser without coding or system setup, and observe their effects on client and global models through real-time visualizations, gaining intuition for challenges such as non-IID data, local overfitting, and scalability. The playground serves as an easy to use educational tool, lowering the entry barrier for newcomers to distributed AI while also offering a sandbox for rapidly prototyping and comparing FL methods. By democratizing exploration of FL, it promotes broader understanding and adoption of this important paradigm.

Federated Learning Playground

TL;DR

Federated Learning Playground is an interactive browser-based platform inspired by and extends TensorFlow Playground that teaches core Federated Learning concepts, lowering the entry barrier for newcomers to distributed AI while also offering a sandbox for rapidly prototyping and comparing FL methods.

Abstract

We present Federated Learning Playground, an interactive browser-based platform inspired by and extends TensorFlow Playground that teaches core Federated Learning (FL) concepts. Users can experiment with heterogeneous client data distributions, model hyperparameters, and aggregation algorithms directly in the browser without coding or system setup, and observe their effects on client and global models through real-time visualizations, gaining intuition for challenges such as non-IID data, local overfitting, and scalability. The playground serves as an easy to use educational tool, lowering the entry barrier for newcomers to distributed AI while also offering a sandbox for rapidly prototyping and comparing FL methods. By democratizing exploration of FL, it promotes broader understanding and adoption of this important paradigm.
Paper Structure (8 sections, 3 figures)

This paper contains 8 sections, 3 figures.

Table of Contents

  1. Introduction
  2. System Design and Implementation
  3. Weight representation.
  4. Round orchestration.
  5. Clustering.
  6. User Interface and Visualizations
  7. Reproducibility and Artifact
  8. Conclusions and Future Work

Figures (3)

  • Figure 1: Client Data Distributions. Top: IID distribution across clients (post-training). Middle: Non-IID distribution (pre-training). Bottom: Uniform class distribution with skewed sample sizes, after local training of clients 0 and 4.
  • Figure 2: Overview of interface controls. FL-specific controls in Rows 2–4, with original controls retained in Row 1.
  • Figure 3: Simulated increase in client dropout halfway through training, with visualizations of Client participation, Comms cost(scales with no. of clients), Client loss distribution(max, mean, min), and Convergence rate.