Memformer: A Memory-Augmented Transformer for Sequence Modeling

Qingyang Wu, Zhenzhong Lan, Kun Qian, Jing Gu, Alborz Geramifard, Zhou Yu

Abstract
Transformers have reached remarkable success in sequence modeling. However, these models have efficiency issues as they need to store all the history token-level representations as memory. We present Memformer, an efficient neural network for sequence modeling, that utilizes an external dynamic memory to encode and retrieve past information. Our model achieves linear time complexity and constant memory space complexity when processing long sequences. We also propose a new optimization scheme, memory replay back-propagation (MRBP), which promotes long-range back-propagation through time with a significantly reduced memory requirement. Experimental results show that Memformer has achieved comparable performance compared against the baselines by using 8.1x less memory space and 3.2x faster on inference. Analysis of the attention pattern shows that our external memory slots can encode and retain important information through timesteps.
Anthology ID:
2022.findings-aacl.29
Volume:
Findings of the Association for Computational Linguistics: AACL-IJCNLP 2022
Month:
November
Year:
2022
Address:
Online only
Editors:
Yulan He, Heng Ji, Sujian Li, Yang Liu, Chua-Hui Chang
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
308–318
Language:
URL:
https://aclanthology.org/2022.findings-aacl.29
DOI:
Bibkey:
Cite (ACL):
Qingyang Wu, Zhenzhong Lan, Kun Qian, Jing Gu, Alborz Geramifard, and Zhou Yu. 2022. Memformer: A Memory-Augmented Transformer for Sequence Modeling. In Findings of the Association for Computational Linguistics: AACL-IJCNLP 2022, pages 308–318, Online only. Association for Computational Linguistics.
Cite (Informal):
Memformer: A Memory-Augmented Transformer for Sequence Modeling (Wu et al., Findings 2022)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-3/2022.findings-aacl.29.pdf
PDF Search