Sequence To Sequence Model

Uses two coupled RNNs in the form of an Encoder and Decoder.

Intuition The encoder serves as some kind of memory that the decoder can rely on.

![[CleanShot 2023-10-03 at 22.11.30@2x.png]]

Example setup with LSTMs for translation.

![[CleanShot 2023-10-03 at 22.11.43@2x.png]]