Structured pruning is a widely used technique for reducing the size of pre-trained language models (PLMs), but current methods often overlook the potential of compressing the hidden dimension (d) in PLMs, a dimension critical to model size and efficiency. This paper introduces a novel structured pruning approach, Structured Pruning with PCA Projection (SP3), targeting the effective reduction of d by projecting features into a space defined by principal components before masking. Extensive experiments on benchmarks (GLUE and SQuAD) show that SP3 can reduce d by 70%, compress 94% of the BERTbase model, maintain over 96% accuracy, and outperform other methods that compress d by 6% in accuracy at the same compression ratio. SP3 has also proven effective with other models, including OPT and Llama. Our data and code are available at an anonymous repo.
@article{arxiv.2308.16475,
title = {$\rm SP^3$: Enhancing Structured Pruning via PCA Projection},
author = {Yuxuan Hu and Jing Zhang and Zhe Zhao and Chen Zhao and Xiaodong Chen and Cuiping Li and Hong Chen},
journal= {arXiv preprint arXiv:2308.16475},
year = {2024}
}