Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT

Jing Zhao, Yifan Wang, Junwei Bao, Youzheng Wu, Xiaodong He


Abstract
Transformer-based pre-trained models, such as BERT, have shown extraordinary success in achieving state-of-the-art results in many natural language processing applications. However, deploying these models can be prohibitively costly, as the standard self-attention mechanism of the Transformer suffers from quadratic computational cost in the input sequence length. To confront this, we propose FCA, a fine- and coarse-granularity hybrid self-attention that reduces the computation cost through progressively shortening the computational sequence length in self-attention. Specifically, FCA conducts an attention-based scoring strategy to determine the informativeness of tokens at each layer. Then, the informative tokens serve as the fine-granularity computing units in self-attention and the uninformative tokens are replaced with one or several clusters as the coarse-granularity computing units in self-attention. Experiments on the standard GLUE benchmark show that BERT with FCA achieves 2x reduction in FLOPs over original BERT with <1% loss in accuracy. We show that FCA offers a significantly better trade-off between accuracy and FLOPs compared to prior methods.
Anthology ID:
2022.acl-long.330
Volume:
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
May
Year:
2022
Address:
Dublin, Ireland
Editors:
Smaranda Muresan, Preslav Nakov, Aline Villavicencio
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
4811–4820
Language:
URL:
https://aclanthology.org/2022.acl-long.330
DOI:
10.18653/v1/2022.acl-long.330
Bibkey:
Cite (ACL):
Jing Zhao, Yifan Wang, Junwei Bao, Youzheng Wu, and Xiaodong He. 2022. Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4811–4820, Dublin, Ireland. Association for Computational Linguistics.
Cite (Informal):
Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT (Zhao et al., ACL 2022)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-3/2022.acl-long.330.pdf
Software:
 2022.acl-long.330.software.tgz
Code
 pierre-zhao/fca-bert
Data
GLUEQNLIRACE