EXCEPTION_MESSAGE_WRONG_RETURN_TYPE_FIT = """
  NumPyClient.fit did not return a tuple with 3 elements.
  The returned values should have the following type signature:
  
      Tuple[NDArrays, int, Dict[str, Scalar]]
  
  Example
  -------
  
      model.get_weights(), 10, {"accuracy": 0.95}
  
  """

def _fit(self: Client, ins: FitIns) -> FitRes:
    """Refine the provided parameters using the locally held dataset."""
    # Deconstruct FitIns
    parameters: NDArrays = parameters_to_ndarrays(ins.parameters)

    # Train
    results = self.numpy_client.fit(parameters, ins.config)  # type: ignore
    if not (
        len(results) == 3
        and isinstance(results[0], list)
        and isinstance(results[1], int)
        and isinstance(results[2], dict)
    ):
        print(results)
        print(len(results))
        
        raise TypeError(EXCEPTION_MESSAGE_WRONG_RETURN_TYPE_FIT)

    class MnistClient(fl.client.NumPyClient):
        def __init__(self, model, X_train) -> None:
            super().__init__()
            self.model = model
            self.X_train = X_train

        def get_parameters(self, config):  # type: ignore
            return utils.get_model_parameters(self.model)

        def fit(self, parameters, config):  # type: ignore
            utils.set_model_params(model, parameters)
            # Ignore convergence failure due to low local epochs
            with warnings.catch_warnings():
                warnings.simplefilter("ignore")
                self.model.fit(self.X_train, y_train)
            print(f"Training finished for round {config['server_round']}")

            params = model.coef_
            # instead of: params = [model.coef_]

            return params, len(X_train), {}

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