Flower Clients in C++ (under development)¶
[!WARNING]
This example is compatible withflwr<1.13.0. We are currently updating it to the newerflwr runway of running Flower Apps.
In this example you will train a linear model on synthetic data using C++ clients.
Acknowledgements¶
Many thanks to the original contributors to this code:
Lekang Jiang (original author and main contributor)
Francisco José SolÃs (code re-organization)
Andreea Zaharia (training algorithm and data generation)
Install requirements¶
You’ll need CMake and Python with flwr installed.
Building the example¶
This example provides you with a CMakeLists.txt file to configure and build the client. Feel free to take a look inside it to see what is happening under the hood.
cmake -S . -B build
cmake --build build
Run the Flower SuperLink, the two clients, and the Flower ServerApp in separate terminals¶
flwr-superlink --insecure
build/flwr_client 0 127.0.0.1:9092
build/flwr_client 1 127.0.0.1:9092
flower-server-app server:app --insecure
