如何在Tensorflow自定义层中初始化输出张量的无形状？

import tensorflow as tf

class CustomLayer(tf.keras.layers.Layer):
    def __init__(self, units):
        super(CustomLayer, self).__init__()
        self.units = units

    def build(self, input_shape):
        self.w = self.add_weight(shape=(input_shape[-1], self.units),
                                 initializer='random_normal',
                                 trainable=True)

    def call(self, inputs):
        output = tf.matmul(inputs, self.w)
        return output

inputs = tf.keras.Input(shape=(10,))
custom_layer = CustomLayer(units=5)
outputs = custom_layer(inputs)

model = tf.keras.Model(inputs=inputs, outputs=outputs)