C5 W4 Ex5- Encoder

I also got to the ‘ListWrapper’ error and understood the correct syntax from the discussions and the tensorflow tutorial.

My question is-- is that a necessary implementation-- storing the encoder layers stacked like that?
or can it be implemented by repeatedly calling the encoder layer with the most resent output?


The stack of layers is an alternative to using inline code for each layer.
You would not get the same results by calling a single encoder in a loop.