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

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

17 Mar 2022ACL 2022 5arXiv:2203.09055archive 2025-07-28

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

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 GLUE and RACE datasets 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.

Paper PDF Conference PDF Code

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

pierre-zhao/fca-bert officialmentioned in papermentioned on GitHubtf report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Informativeness

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Absolute Position Encodings Adam Attention Attention Dropout BERT BPE Dense Connections Dropout Label Smoothing Layer Normalization Linear Layer Linear Warmup With Linear Decay Multi-Head Attention Position-Wise Feed-Forward Layer Residual Connection Softmax Transformer Weight Decay WordPiece

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections