Multi-Head Attention
Running several attention operations in parallel over different projections.
Resumen
Multi-head attention splits the representation into several subspaces, runs attention independently in each, and concatenates the results. It solves the problem that a single attention distribution has to average over incompatible kinds of relationship - syntactic, semantic, positional - in one weighting. It exists because several narrow attentions outperform one wide one at the same cost.
Grafo de tecnologías
Cómo se conecta Multi-Head Attention con el resto del stack.
Classification
- Transformer architecture Algorithms Multi-Head Attention part of Transformer High ✓ source
Herramientas
Software que implementa o se integra con Multi-Head Attention.