LLM-Rank: A Graph Theoretical Approach to Pruning Large Language Models
This addresses the deployment cost issue for users of large language models, though it is an incremental improvement over existing pruning techniques.
The paper tackles the problem of high computational and memory costs in large language models by proposing a novel pruning method based on graph theory centrality measures, achieving 6.09% higher accuracy retention for MLPRank and 13.42% for LLMRank compared to baselines.
The evolving capabilities of large language models are accompanied by growing sizes and deployment costs, necessitating effective inference optimisation techniques. We propose a novel pruning method utilising centrality measures from graph theory, reducing both the computational requirements and the memory footprint of these models. Specifically, we devise a method for creating a weighted directed acyclical graph representation of multilayer perceptrons to which we apply a modified version of the weighted PageRank centrality measure to compute node importance scores. In combination with uniform pruning this leads to structured sparsity. We call this pruning method MLPRank. Furthermore we introduce an extension to decoder-only transformer models and call it LLMRank. For both variants we demonstrate a strong performance. With MLPRank on average leading to 6.09 % higher accuracy retention than three popular baselines and 13.42 % with LLMRank compared to two popular baselines. Code is available at https://github.com/amazon-science/llm-rank-pruning.