One-shot Sensitivity-aware Mixed Sparsity Pruning For Large Language Models

Shao Hang, Liu Bei, Xiao Bo, Zeng Ke, Wan Guanglu, Qian Yanmin. Arxiv 2023

[Paper]    
Applications Efficiency And Optimization GPT Language Modeling Model Architecture Pretraining Methods Pruning Quantization Reinforcement Learning Training Techniques Transformer

Various Large Language Models~(LLMs) from the Generative Pretrained Transformer(GPT) family have achieved outstanding performances in a wide range of text generation tasks. However, the enormous model sizes have hindered their practical use in real-world applications due to high inference latency. Therefore, improving the efficiencies of LLMs through quantization, pruning, and other means has been a key issue in LLM studies. In this work, we propose a method based on Hessian sensitivity-aware mixed sparsity pruning to prune LLMs to at least 50% sparsity without the need of any retraining. It allocates sparsity adaptively based on sensitivity, allowing us to reduce pruning-induced error while maintaining the overall sparsity level. The advantages of the proposed method exhibit even more when the sparsity is extremely high. Furthermore, our method is compatible with quantization, enabling further compression of LLMs. We have released the available code.

Similar Work