Chapter 13 - Machine Translation Page 257, Figure 13.6: Encoder hidden layer indices should be from $1$ to $n$. Output length is not dependent on input length. Index of the last $y$ should be $m$ instead of $n$.
