DuoGPT: Training-free Dual Sparsity through Activation-aware Pruning in LLMs

Ruokai Yin, Yuhang Li, Donghyun Lee, Priyadarshini Panda

Keywords: Weight Pruning, Dual-sparsity, Activation Sparsity, Efficient Large Language Models, Post-training Calibration, Model Compression

TL;DR: We propose DuoGPT, a training-free pruning framework that integrates activation sparsity into OBC framework to enable efficient dual-sparse LLM inference with state-of-the-art accuracy–efficiency trade-offs and scalability.

Abstract: Large language models (LLMs) deliver strong performance but are difficult to deploy due to high memory and compute costs. While pruning reduces these demands, most methods ignore activation sparsity observed at runtime. We reinterpret activation sparsity as dynamic structured weight sparsity and propose DuoGPT, a unified framework that constructs dual-sparse (spMspV) workloads by combining unstructured weight pruning with activation sparsity. To preserve accuracy, we extend the Optimal Brain Compression (OBC) framework with activation-aware calibration and introduce output residuals from the dense model as correction terms. We further optimize the solution for efficient GPU execution, enabling scalability to billion-parameter LLMs. Evaluations on LLaMA-2 and LLaMA-3 show that DuoGPT outperforms state-of-the-art structured pruning methods by up to 9.17\% accuracy at an iso-speedup of 1.39$\times$ compared to the baseline dense model. Code is available at GitHub.

Primary Area: Deep learning (e.g., architectures, generative models, optimization for deep networks, foundation models, LLMs)

Submission Number: 1185