2024/07/26 by Jianwei Li, Li, Jianwei, Yijun Dong +3 · 3 citations
Computer Science · Psychology · #Agronomy #Algorithms and Data Compression #Artificial Intelligence (cs.AI) #Biology #Computer science #Economics #FOS: Computer and information sciences #Natural Language Processing Techniques #Pruning #Psychology #Retraining
paper · pdf · doi:10.48550/arxiv.2407.19126
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2024/07/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
To remove redundant components of large language models (LLMs) without incurring significant computational costs, this work focuses on single-shot pruning without a retraining phase. We simplify the pruning process for Transformer-based LLMs by identifying a depth-2 pruning structure that functions independently. Additionally, we propose two inference-aware pruning criteria derived from the optimization perspective of output approximation, which outperforms traditional training-aware metrics such as gradient and Hessian. We also introduce a two-step reconstruction technique to mitigate pruning errors without model retraining. Experimental results demonstrate that our approach significantly reduces computational costs and hardware requirements while maintaining superior performance across various datasets and models.