shaoxiongji/federated-learning
A PyTorch implementation of the FedAVG federated learning algorithm for training deep neural networks on decentralized data.
Velocity · 7d
+0.5
★ / day
Trend
→steady
star history
Implements federated averaging (FedAVG) for distributed deep learning across multiple clients without centralizing data. Supports MLP and CNN models on MNIST and CIFAR10 datasets with both IID and non-IID data distributions. Reproduces the foundational federated learning paper by McMahan et al.