Talking-Heads Attention

Code

• github.com/zygmuntz/hyperband
  Officialnone★ 596
• github.com/JunnYu/x-transformers-paddle
  jax★ 10
• github.com/tensorflow/models/tree/master/official/nlp/modeling/layers
  tf★ 0
• github.com/lucidrains/x-transformers
  pytorch★ 0

Abstract

We introduce "talking-heads attention" - a variation on multi-head attention which includes linearprojections across the attention-heads dimension, immediately before and after the softmax operation.While inserting only a small number of additional parameters and a moderate amount of additionalcomputation, talking-heads attention leads to better perplexities on masked language modeling tasks, aswell as better quality when transfer-learning to language comprehension and question answering tasks.

Tasks

Reproductions