Démarrage rapide de scikit-learn

In this federated learning tutorial we will learn how to train a Logistic Regression on MNIST using Flower and scikit-learn. It is recommended to create a virtual environment and run everything within a virtualenv.

Let’s use flwr new to create a complete Flower+scikit-learn project. It will generate all the files needed to run, by default with the Flower Simulation Engine, a federation of 10 nodes using FedAvg The dataset will be partitioned using Flower Datasets’s IidPartitioner

Now that we have a rough idea of what this example is about, let’s get started. First, install Flower in your new environment:

# In a new Python environment
$ pip install flwr

Then, run the command below. You will be prompted to select one of the available templates (choose sklearn), give a name to your project, and type in your developer name:

$ flwr new

After running it you’ll notice a new directory with your project name has been created. It should have the following structure:

<your-project-name>
├── <your-project-name>
│   ├── __init__.py
│   ├── client_app.py   # Defines your ClientApp
│   ├── server_app.py   # Defines your ServerApp
│   └── task.py         # Defines your model, training and data loading
├── pyproject.toml      # Project metadata like dependencies and configs
└── README.md

If you haven’t yet installed the project and its dependencies, you can do so by:

# From the directory where your pyproject.toml is
$ pip install -e .

To run the project, do:

# Run with default arguments
$ flwr run .

With default arguments you will see an output like this one:

Loading project configuration...
Success
INFO :      Starting Flower ServerApp, config: num_rounds=3, no round_timeout
INFO :
INFO :      [INIT]
INFO :      Requesting initial parameters from one random client
INFO :      Received initial parameters from one random client
INFO :      Starting evaluation of initial global parameters
INFO :      Evaluation returned no results (`None`)
INFO :
INFO :      [ROUND 1]
INFO :      configure_fit: strategy sampled 10 clients (out of 10)
INFO :      aggregate_fit: received 10 results and 0 failures
WARNING :   No fit_metrics_aggregation_fn provided
INFO :      configure_evaluate: strategy sampled 10 clients (out of 10)
INFO :      aggregate_evaluate: received 10 results and 0 failures
WARNING :   No evaluate_metrics_aggregation_fn provided
INFO :
INFO :      [ROUND 2]
INFO :      configure_fit: strategy sampled 10 clients (out of 10)
INFO :      aggregate_fit: received 10 results and 0 failures
INFO :      configure_evaluate: strategy sampled 10 clients (out of 10)
INFO :      aggregate_evaluate: received 10 results and 0 failures
INFO :
INFO :      [ROUND 3]
INFO :      configure_fit: strategy sampled 10 clients (out of 10)
INFO :      aggregate_fit: received 10 results and 0 failures
INFO :      configure_evaluate: strategy sampled 10 clients (out of 10)
INFO :      aggregate_evaluate: received 10 results and 0 failures
INFO :
INFO :      [SUMMARY]
INFO :      Run finished 3 round(s) in 19.41s
INFO :          History (loss, distributed):
INFO :                  round 1: 1.3447584261018466
INFO :                  round 2: 0.9680018613482815
INFO :                  round 3: 0.7667920399137523
INFO :

You can also override the parameters defined in the [tool.flwr.app.config] section in pyproject.toml like this:

# Override some arguments
$ flwr run . --run-config "num-server-rounds=5 local-epochs=2"

What follows is an explanation of each component in the project you just created: dataset partition, the model, defining the ClientApp and defining the ServerApp.

The Data

This tutorial uses Flower Datasets to easily download and partition the MNIST dataset. In this example you’ll make use of the IidPartitioner to generate num_partitions partitions. You can choose other partitioners available in Flower Datasets. Each ClientApp will call this function to create dataloaders with the data that correspond to their data partition.

partitioner = IidPartitioner(num_partitions=num_partitions)
fds = FederatedDataset(
    dataset="mnist",
    partitioners={"train": partitioner},
)

dataset = fds.load_partition(partition_id, "train").with_format("numpy")

X, y = dataset["image"].reshape((len(dataset), -1)), dataset["label"]

# Split the on edge data: 80% train, 20% test
X_train, X_test = X[: int(0.8 * len(X))], X[int(0.8 * len(X)) :]
y_train, y_test = y[: int(0.8 * len(y))], y[int(0.8 * len(y)) :]

The Model

We define the LogisticRegression model from scikit-learn in the get_model() function:

def get_model(penalty: str, local_epochs: int):

    return LogisticRegression(
        penalty=penalty,
        max_iter=local_epochs,
        warm_start=True,
    )

To perform the training and evaluation, we will make use of the .fit() and .score() methods available in the LogisticRegression class.

The ClientApp

The main changes we have to make to use scikit-learn with Flower will be found in the get_model_params(), set_model_params(), and set_initial_params() functions. In get_model_params(), the coefficients and intercept of the logistic regression model are extracted and represented as a list of NumPy arrays. In set_model_params(), that’s the opposite: given a list of NumPy arrays it applies them to an existing LogisticRegression model. Finally, in set_initial_params(), we initialize the model parameters based on the MNIST dataset, which has 10 classes (corresponding to the 10 digits) and 784 features (corresponding to the size of the MNIST image array, which is 28 × 28). Doing this is fairly easy in scikit-learn.

def get_model_params(model):
    if model.fit_intercept:
        params = [
            model.coef_,
            model.intercept_,
        ]
    else:
        params = [model.coef_]
    return params


def set_model_params(model, params):
    model.coef_ = params[0]
    if model.fit_intercept:
        model.intercept_ = params[1]
    return model


def set_initial_params(model):
    n_classes = 10  # MNIST has 10 classes
    n_features = 784  # Number of features in dataset
    model.classes_ = np.array([i for i in range(10)])

    model.coef_ = np.zeros((n_classes, n_features))
    if model.fit_intercept:
        model.intercept_ = np.zeros((n_classes,))

The rest of the functionality is directly inspired by the centralized case:

class FlowerClient(NumPyClient):
    def __init__(self, model, X_train, X_test, y_train, y_test):
        self.model = model
        self.X_train = X_train
        self.X_test = X_test
        self.y_train = y_train
        self.y_test = y_test

    def fit(self, parameters, config):
        set_model_params(self.model, parameters)

        # Ignore convergence failure due to low local epochs
        with warnings.catch_warnings():
            warnings.simplefilter("ignore")
            self.model.fit(self.X_train, self.y_train)

        return get_model_params(self.model), len(self.X_train), {}

    def evaluate(self, parameters, config):
        set_model_params(self.model, parameters)
        loss = log_loss(self.y_test, self.model.predict_proba(self.X_test))
        accuracy = self.model.score(self.X_test, self.y_test)
        return loss, len(self.X_test), {"accuracy": accuracy}

Finally, we can construct a ClientApp using the FlowerClient defined above by means of a client_fn() callback. Note that the context enables you to get access to hyperparemeters defined in your pyproject.toml to configure the run. In this tutorial we access the local-epochs setting to control the number of epochs a ClientApp will perform when running the fit() method. You could define additioinal hyperparameters in pyproject.toml and access them here.

def client_fn(context: Context):
    # Load data and model
    partition_id = context.node_config["partition-id"]
    num_partitions = context.node_config["num-partitions"]
    X_train, X_test, y_train, y_test = load_data(partition_id, num_partitions)
    penalty = context.run_config["penalty"]
    local_epochs = context.run_config["local-epochs"]
    model = get_model(penalty, local_epochs)

    # Setting initial parameters, akin to model.compile for keras models
    set_initial_params(model)

    # Return Client instance
    return FlowerClient(model, X_train, X_test, y_train, y_test).to_client()


# Flower ClientApp
app = ClientApp(client_fn)

The ServerApp

To construct a ServerApp we define a server_fn() callback with an identical signature to that of client_fn() but the return type is ServerAppComponents as opposed to a Client In this example we use the FedAvg strategy. To it we pass a zero-initialized model that will server as the global model to be federated. Note that the values of num-server-rounds, penalty, and local-epochs are read from the run config. You can find the default values defined in the pyproject.toml.

def server_fn(context: Context):
    # Read from config
    num_rounds = context.run_config["num-server-rounds"]

    # Create LogisticRegression Model
    penalty = context.run_config["penalty"]
    local_epochs = context.run_config["local-epochs"]
    model = get_model(penalty, local_epochs)

    # Setting initial parameters, akin to model.compile for keras models
    set_initial_params(model)

    initial_parameters = ndarrays_to_parameters(get_model_params(model))

    # Define strategy
    strategy = FedAvg(
        fraction_fit=1.0,
        fraction_evaluate=1.0,
        min_available_clients=2,
        initial_parameters=initial_parameters,
    )
    config = ServerConfig(num_rounds=num_rounds)

    return ServerAppComponents(strategy=strategy, config=config)


# Create ServerApp
app = ServerApp(server_fn=server_fn)

Congratulations! You’ve successfully built and run your first federated learning system in scikit-learn.

Note

Check the source code of the extended version of this tutorial in examples/sklearn-logreg-mnist in the Flower GitHub repository.