concision_neural/layers/attention/
multi_head.rs

1/*
2    Appellation: multi-head <module>
3    Contrib: @FL03
4*/
5
6/// Multi-Headed attention is the first evolution of the Scaled Dot-Product Attention
7/// mechanism. They allow the model to jointly attend to information from different
8/// representation subspaces at different positions.
9pub struct MultiHeadAttention;