Efficient Long Sequence Modeling via State Space Augmented Transformer

Transformer models have achieved superior performance in various natural language processing tasks. However, the quadratic computational cost of the attention mechanism limits its practicality for long sequences. There are existing attention variants that improve the computational efficiency, but th...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2022-12
Main Authors: Zuo, Simiao, Liu, Xiaodong, Jiao, Jian, Denis, Charles, Manavoglu, Eren, Zhao, Tuo, Gao, Jianfeng
Format: Article
Language:English
Subjects:
Coders
Computational efficiency
Computing costs
Encoders-Decoders
Modelling
Natural language
Natural language processing
Speech recognition
State space models
Transformers
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Transformer models have achieved superior performance in various natural language processing tasks. However, the quadratic computational cost of the attention mechanism limits its practicality for long sequences. There are existing attention variants that improve the computational efficiency, but they have limited ability to effectively compute global information. In parallel to Transformer models, state space models (SSMs) are tailored for long sequences, but they are not flexible enough to capture complicated local information. We propose SPADE, short for \(\underline{\textbf{S}}\)tate s\(\underline{\textbf{P}}\)ace \(\underline{\textbf{A}}\)ugmente\(\underline{\textbf{D}}\) Transform\(\underline{\textbf{E}}\)r. Specifically, we augment a SSM into the bottom layer of SPADE, and we employ efficient local attention methods for the other layers. The SSM augments global information, which complements the lack of long-range dependency issue in local attention methods. Experimental results on the Long Range Arena benchmark and language modeling tasks demonstrate the effectiveness of the proposed method. To further demonstrate the scalability of SPADE, we pre-train large encoder-decoder models and present fine-tuning results on natural language understanding and natural language generation tasks.
ISSN:2331-8422