class GRU:
    def __init__(self, input_size, hidden_size, output_size, init_method='xavier'):
        self.input_size = input_size
        self.hidden_size = hidden_size
        self.output_size = output_size
        self.weight_initializer = WeightInitializer(method=init_method)

        # Initialize weights
        self.wz = self.weight_initializer.initialize((hidden_size, hidden_size + input_size))
        self.wr = self.weight_initializer.initialize((hidden_size, hidden_size + input_size))
        self.wh = self.weight_initializer.initialize((hidden_size, hidden_size + input_size))

        # Initialize biases
        self.bz = np.zeros((hidden_size, 1))
        self.br = np.zeros((hidden_size, 1))
        self.bh = np.zeros((hidden_size, 1))

        # Initialize output layer weights and biases
        self.why = self.weight_initializer.initialize((output_size, hidden_size))
        self.by = np.zeros((output_size, 1))