如何将多个参数从Tensorflow probability传递到DistributionLambda层？

import tensorflow as tf
import tensorflow_probability as tfp

class CustomDistributionLayer(tfp.layers.DistributionLambda):
    def __init__(self, **kwargs):
        super().__init__(**kwargs)

    def call(self, inputs):
        # 解包输入参数
        param1, param2, param3 = inputs

        # 创建分布对象
        distribution = tfp.distributions.Normal(loc=param1, scale=param2)

        return distribution

model = tf.keras.Sequential([
    # 其他层...
    CustomDistributionLayer(name='custom_distribution'),
    # 其他层...
])