FedOBP

Official repository for FedOBP

This repository is based on the FL-bench implementation.

Installation

pip install -r .env/requirements.txt

Step 1. Generate FL Dataset

Partition the MNIST according to Dir(0.1) for 100 clients

python generate_data.py -d mnist -a 0.1 -cn 100

About methods of generating federated dastaset, go check data/README.md for full details.

Step 2. Run FedOBP Main Experiment

python main_fedobp.py [--config-path, --config-name] [dataset.name=<DATASET_NAME> args...]

Step 4. FedOBP Ablation Experiment

python run_script_ablation.py

Step 3. Run Baselines Experiment

python main.py [--config-path, --config-name] [method=<METHOD_NAME> args...]

Monitor runs

This implementation supports tensorboard.

1. Run tensorboard --logdir=<your_log_dir> on terminal.
2. Go check localhost:6006 on your browser.

Bibtex

@inproceedings{chen2025fedobp,
  title={FedOBP: Federated Optimal Brain Personalization with Few Personalized Parameters},
  author={Chen, Xingyan and Du, Tian and Diao, Enmao},
  year={2025},
  url={https://github.com/uglyghost/FedOBP.git}
}