Tutorials

A variety of resources are available to help you learn Federated Learning (FL) programming with NVFLARE:

Video Tutorials

• Quick introductions to FLARE for deep learning and traditional machine learning.

• Examples include CIFAR10 classification and Kaplan-Meier survival analysis.

• Watch videos and see code

Tutorial Catalog

• Browse and filter a comprehensive set of examples by experience level, framework, algorithm, application, industry, API, and privacy algorithm.

• Explore the Catalog

Step-by-Step Walkthroughs

• Explore different FL algorithms using the same datasets (CIFAR10 for images, Higgs for tabular data).

• Covers FedAvg, Cyclic, Swarm Learning, Federated Statistics, Scikit-learn, XGBoost, and more.

• Step-by-step examples

ML-to-FL Conversion

• Learn to convert standalone or centralized training code to FL code with various deep learning frameworks.

• ML-to-FL walkthroughs

Feature Tutorials

• POC Tutorials:

• Simulator CLI & Python API

• FLARE Python API: Job Submission & Monitoring

• Logging: Configuration & Customization

• Job CLI: Job Submission & Templates

Self-Paced Learning

• Extensive materials for self-paced training with NVIDIA FLARE, including detailed tutorials and resources.

• For a comprehensive guide, see the Self-Paced-Training Tutorials documentation.

• 12-chapter course: in-depth overview of FLARE, covering running federated learning applications, algorithms, system architecture, experimental tracking, system monitoring, and industrial applications.

• Each notebook can be run independently, but sequential learning is recommended.

• Over 100 notebooks and 80 videos: a thorough guide to federated learning with FLARE.