Use the mask_tensor inside the example code.

Author: saberkun

rfcs/20200616-keras-multihead-attention.md

@@ -134,7 +134,7 @@ class MultiHeadAttention(tf.keras.layers.Layer):
   >>> target = tf.keras.Input(shape=[8, 16])
   >>> source = tf.keras.Input(shape=[4, 16])
   >>> mask_tensor = tf.keras.Input(shape=[8, 4])
-  >>> output_tensor, weights = layer([target, source])
+  >>> output_tensor, weights = layer([target, source], attention_mask=mask_tensor)
   >>> print(output_tensor.shape), print(weights.shape)
   (None, 8, 16)  (None, 2, 8, 4)