Examples

Complete, worked walkthroughs: take a real dataset and model, and run it as a federated learning project on BranchKey — step by step, from the ground up. Each example is self-contained; read it in order and adapt the code to your own setup.

These guides build on the Concepts section — especially Extracting & reloading weights — so it helps to have read those first.

Available examples

🧠 Brain metastasis segmentation (UCSF-BMSR)

Segment brain metastases on MRI using the UCSF-BMSR dataset (T1-weighted post-contrast). You build a small 2D U-Net from scratch in PyTorch — where extracting and reloading weights is trivial — then federate it across sites with BranchKey. A deliberate stepping stone before heavier frameworks like nnU-Net.

More examples will be added here over time.