xgboost/demo/nvflare/README.md

Experimental Support of Federated XGBoost using NVFlare

This directory contains a demo of Federated Learning using NVFlare.

Training with CPU only

To run the demo, first build XGBoost with the federated learning plugin enabled (see the README).

Install NVFlare (note that currently NVFlare only supports Python 3.8):

pip install nvflare

Prepare the data:

./prepare_data.sh

Start the NVFlare federated server:

/tmp/nvflare/poc/server/startup/start.sh

In another terminal, start the first worker:

/tmp/nvflare/poc/site-1/startup/start.sh

And the second worker:

/tmp/nvflare/poc/site-2/startup/start.sh

Then start the admin CLI:

/tmp/nvflare/poc/admin/startup/fl_admin.sh

In the admin CLI, run the following command:

submit_job hello-xgboost

Once the training finishes, the model file should be written into /tmp/nvlfare/poc/site-1/run_1/test.model.json and /tmp/nvflare/poc/site-2/run_1/test.model.json respectively.

Finally, shutdown everything from the admin CLI, using admin as password:

shutdown client
shutdown server

Training with GPUs

To demo with Federated Learning using GPUs, make sure your machine has at least 2 GPUs. Build XGBoost with the federated learning plugin enabled along with CUDA, but with NCCL turned off (see the README).

Modify config/config_fed_client.json and set use_gpus to true, then repeat the steps above.