With the hype around AI not likely to slow down anytime soon, it’s time to give transformers their due, which is why I’d like to explain ... model follows an encoder-decoder architecture.