README.md

NVFlare hello-world examples

This folder contains hello-world examples for NVFlare.

Please make sure you set up a virtual environment and install JupyterLab following the example root readme.

Please also install "./requirements.txt" in each example folder.

Hello World Notebook

Prerequisites

Before you run the notebook, the following preparation work must be done:

1. Install a virtual environment following the instructions in example root readme
2. Install Jupyter Lab and install a new kernel for the virtualenv called nvflare_example
3. Run hw_pre_start.sh in the terminal before running the notebook
4. Run hw_post_cleanup.sh in the terminal after running the notebook

• Hello world notebook

Hello World Examples

Easier ML/DL to FL transition

• ML to FL: Showcases how to convert existing ML/DL code to an NVFlare job.

Step by step examples

• Step by step examples: Shows specific techniques and workflows and what needs to be changed for each.

Workflows

• Hello Scatter and Gather
  • Example using ScatterAndGather controller workflow.
• Hello Cross-Site Validation
  • Example using CrossSiteModelEval controller workflow.
• Hello Cyclic Weight Transfer
  • Example using CyclicController controller workflow to implement Cyclic Weight Transfer.
• Hello Client Controlled Workflows
  • Example using Client Controlled Workflows.

Deep Learning

• Hello PyTorch
  • Example using NVIDIA FLARE an image classifier using (FedAvg) and PyTorch as the deep learning training framework.
• Hello TensorFlow
  • Example of using NVIDIA FLARE an image classifier using (FedAvg) and TensorFlow as the deep learning training framework.