Hi! Great post, helped me a lot!

I have only one question in this comment: “# mask the encoder outputs for the 2nd attention block” ..

So it is passing “inputs” in “tf.keras.layers.Lambda”.

For the comment, would it be “enc_outputs”?