Drizzlingg/FedLab-benchmarks

Standard federated learning implementations in FedLab and FL benchmarks.

FedLab-benchmarks

This repo contains standard FL algorithm implementations and FL benchmarks using FedLab.

Currently, following algorithms or benchrmarks are available：

Optimization Algorithms

• FedAvg: Communication-Efficient Learning of Deep Networks from Decentralized Data
• FedAsync: Asynchronous Federated Optimization
• FedProx: Federated Optimization in Heterogeneous Networks
• FedDyn: Federated Learning based on Dynamic Regularization
• Personalized-FedAvg: Improving Federated Learning Personalization via Model Agnostic Meta Learning

Compression Algorithms

• DGC: Deep Gradient Compression: Reducing the Communication Bandwidth for Distributed Training
• QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding

Datasets

• LEAF: A Benchmark for Federated Settings
• NIID-Bench: Federated Learning on Non-IID Data Silos: An Experimental Study

Working list

• PFL: Debiasing Model Updates for Improving Personalized Federated Training
• qFFL: Fair Resource Allocation in Federated Learning
• FedMGDA+: Federated Learning meets Multi-objective Optimization

---

We highly welcome you to contribute federated learning algorithm based on FedLab. If you encounter any problems, do not hesitate to submit an issue or send an email to repo maintainers.