The encoder
The decoder
The position embedding
The encoder block
The self-attention layer
The layer-normalization
The position-wise feed-forward network
The decoder block
The cross-attention layer
The predicting head
The overall architecture
The architecture is composed of an encoder and a decoder.
