解决 d2l 动手学深度学习 7.5. 批量规范化 AttributeError: ‘BatchNorm‘ object has no attribute ‘add_update‘

李沐本书项目中提供的这一节的 TensorFlow 版 ipynb 文件中的代码：

class BatchNorm(tf.keras.layers.Layer):
    def __init__(self, **kwargs):
        super(BatchNorm, self).__init__(**kwargs)

    def build(self, input_shape):
        weight_shape = [input_shape[-1], ]
        # 参与求梯度和迭代的拉伸和偏移参数，分别初始化成1和0
        self.gamma = self.add_weight(name='gamma', shape=weight_shape,
            initializer=tf.initializers.ones, trainable=True)
        self.beta = self.add_weight(name='beta', shape=weight_shape,
            initializer=tf.initializers.zeros, trainable=True)
        # 非模型参数的变量初始化为0和1
        self.moving_mean = self.add_weight(name='moving_mean',
            shape=weight_shape, initializer=tf.initializers.zeros,
            trainable=False)
        self.moving_variance = self.add_weight(name='moving_variance',
            shape=weight_shape, initializer=tf.initializers.ones,