Luna: Linear Unified Nested Attention

by   Xuezhe Ma, et al.

The quadratic computational and memory complexities of the Transformer's attention mechanism have limited its scalability for modeling long sequences. In this paper, we propose Luna, a linear unified nested attention mechanism that approximates softmax attention with two nested linear attention functions, yielding only linear (as opposed to quadratic) time and space complexity. Specifically, with the first attention function, Luna packs the input sequence into a sequence of fixed length. Then, the packed sequence is unpacked using the second attention function. As compared to a more traditional attention mechanism, Luna introduces an additional sequence with a fixed length as input and an additional corresponding output, which allows Luna to perform attention operation linearly, while also storing adequate contextual information. We perform extensive evaluations on three benchmarks of sequence modeling tasks: long-context sequence modeling, neural machine translation and masked language modeling for large-scale pretraining. Competitive or even better experimental results demonstrate both the effectiveness and efficiency of Luna compared to a variety


page 1

page 2

page 3

page 4


Adaptive Multi-Resolution Attention with Linear Complexity

Transformers have improved the state-of-the-art across numerous tasks in...

Random Feature Attention

Transformers are state-of-the-art models for a variety of sequence model...

Agglomerative Attention

Neural networks using transformer-based architectures have recently demo...

ABC: Attention with Bounded-memory Control

Transformer architectures have achieved state-of-the-art results on a va...

A Cheap Linear Attention Mechanism with Fast Lookups and Fixed-Size Representations

The softmax content-based attention mechanism has proven to be very bene...

ERNIE-DOC: The Retrospective Long-Document Modeling Transformer

Transformers are not suited for processing long document input due to it...

Linearizing Transformer with Key-Value Memory Bank

Transformer has brought great success to a wide range of natural languag...

Code Repositories


A PyTorch Implementation of the Luna: Linear Unified Nested Attention

view repo