Secure aggregation with Flower (the SecAgg+ protocol)ΒΆ
The following steps describe how to use Flowerβs built-in Secure Aggregation components. This example demonstrates how to apply SecAgg+
to the same federated learning workload as in the quickstart-pytorch example. The ServerApp
uses the SecAggPlusWorkflow
while ClientApp
uses the secaggplus_mod
. To introduce the various steps involved in SecAgg+
, this example introduces as a sub-class of SecAggPlusWorkflow
the SecAggPlusWorkflowWithLogs
. It is enabled by default, but you can disable (see later in this readme).
Set up the projectΒΆ
Clone the projectΒΆ
Start by cloning the example project:
git clone --depth=1 https://github.com/adap/flower.git _tmp \
&& mv _tmp/examples/flower-secure-aggregation . \
&& rm -rf _tmp && cd flower-secure-aggregation
This will create a new directory called flower-secure-aggregation
containing the
following files:
flower-secure-aggregation
|
βββ secaggexample
| βββ __init__.py
| βββ client_app.py # Defines your ClientApp
| βββ server_app.py # Defines your ServerApp
| βββ task.py # Defines your model, training and data loading
| βββ workflow_with_log.py # Defines a workflow used when `is-demo=true`
βββ pyproject.toml # Project metadata like dependencies and configs
βββ README.md
Install dependencies and projectΒΆ
Install the dependencies defined in pyproject.toml
as well as the secaggexample
package.
pip install -e .
Run the projectΒΆ
You can run your Flower project in both simulation and deployment mode without making changes to the code. If you are starting with Flower, we recommend you using the simulation mode as it requires fewer components to be launched manually. By default, flwr run
will make use of the Simulation Engine.
Run with the Simulation EngineΒΆ
[!NOTE] Check the Simulation Engine documentation to learn more about Flower simulations and how to optimize them.
flwr run .
You can also override some of the settings for your ClientApp
and ServerApp
defined in pyproject.toml
. For example
flwr run . --run-config "num-server-rounds=5 learning-rate=0.25"
To adapt the example for a practial usage, set is-demo=false
like shown below. You might want to adjust the num-shares
and reconstruction-threshold
settings to suit your requirements. You can override those via --run-config
as well.
flwr run . --run-config is-demo=false
Run with the Deployment EngineΒΆ
Follow this how-to guide to run the same app in this example but with Flowerβs Deployment Engine. After that, you might be intersted in setting up secure TLS-enabled communications and SuperNode authentication in your federation.
If you are already familiar with how the Deployment Engine works, you may want to learn how to run it using Docker. Check out the Flower with Docker documentation.