Once is Enough: A Light-Weight Cross-Attention for Fast Sentence Pair Modeling

TY - UNPB

T1 - Once is Enough

T2 - A Light-Weight Cross-Attention for Fast Sentence Pair Modeling

AU - Yang, Yuanhang

AU - qi, shiyi

AU - Gao, Cuiyun

AU - Xu, Zenglin

AU - He, Yulan

AU - Wang, Qifan

AU - Liu, Chuanyi

PY - 2022/10/11

Y1 - 2022/10/11

N2 - Transformer-based models have achieved great success on sentence pair modeling tasks, such as answer selection and natural language inference (NLI). These models generally perform cross-attention over input pairs, leading to prohibitive computational costs. Recent studies propose dual-encoder and late interaction architectures for faster computation. However, the balance between the expressive of cross-attention and computation speedup still needs better coordinated. To this end, this paper introduces a novel paradigm MixEncoder for efficient sentence pair modeling. MixEncoder involves a light-weight cross-attention mechanism. It conducts query encoding only once while modeling the query-candidate interaction in parallel. Extensive experiments conducted on four tasks demonstrate that our MixEncoder can speed up sentence pairing by over 113x while achieving comparable performance as the more expensive cross-attention models.

AB - Transformer-based models have achieved great success on sentence pair modeling tasks, such as answer selection and natural language inference (NLI). These models generally perform cross-attention over input pairs, leading to prohibitive computational costs. Recent studies propose dual-encoder and late interaction architectures for faster computation. However, the balance between the expressive of cross-attention and computation speedup still needs better coordinated. To this end, this paper introduces a novel paradigm MixEncoder for efficient sentence pair modeling. MixEncoder involves a light-weight cross-attention mechanism. It conducts query encoding only once while modeling the query-candidate interaction in parallel. Extensive experiments conducted on four tasks demonstrate that our MixEncoder can speed up sentence pairing by over 113x while achieving comparable performance as the more expensive cross-attention models.

KW - cs.CL

KW - cs.AI

M3 - Preprint

BT - Once is Enough

ER -