Efficient Transformers: A Survey

TL;DR

This survey reviews recent efficient Transformer architectures designed to reduce computational and memory costs, providing an organized overview of models like Reformer, Linformer, Performer, and Longformer across various domains.

Contribution

It offers a comprehensive and organized overview of recent efficiency-focused Transformer models, aiding researchers in understanding current advancements and trends.

Findings

Survey covers a wide range of efficiency improvements

Highlights key models and their domain applications

Provides insights into future research directions

Abstract

Transformer model architectures have garnered immense interest lately due to their effectiveness across a range of domains like language, vision and reinforcement learning. In the field of natural language processing for example, Transformers have become an indispensable staple in the modern deep learning stack. Recently, a dizzying number of "X-former" models have been proposed - Reformer, Linformer, Performer, Longformer, to name a few - which improve upon the original Transformer architecture, many of which make improvements around computational and memory efficiency. With the aim of helping the avid researcher navigate this flurry, this paper characterizes a large and thoughtful selection of recent efficiency-flavored "X-former" models, providing an organized and comprehensive overview of existing work and models across multiple domains.