ECoFLaP: Efficient Coarse-to-Fine Layer-Wise Pruning for Vision-Language Models

TL;DR

ECoFLaP is a two-stage pruning method for large vision-language models that efficiently reduces model size by combining global importance scoring with layer-wise pruning, maintaining high performance.

Contribution

It introduces a global importance-based sparsity determination followed by layer-wise pruning, improving compression efficiency and performance over existing methods.

Findings

Significant performance gains at high sparsity levels.

Effective model compression across various multimodal datasets.

Outperforms prevalent pruning techniques in experiments.

Abstract

Large Vision-Language Models (LVLMs) can understand the world comprehensively by integrating rich information from different modalities, achieving remarkable advancements on various multimodal downstream tasks. However, deploying LVLMs is often problematic due to their massive computational/energy costs and carbon consumption. Such issues make it infeasible to adopt conventional iterative global pruning, which is costly due to computing the Hessian matrix of the entire large model for sparsification. Alternatively, several studies have recently proposed layer-wise pruning approaches to avoid the expensive computation of global pruning and efficiently compress model weights according to their importance within a layer. However, they often suffer from suboptimal model compression due to their lack of a global perspective. To address this limitation in recent efficient pruning methods for large models, we propose Efficient Coarse-to-Fine LayerWise Pruning (ECoFLaP), a two-stage coarse-to-fine weight pruning approach for LVLMs. We first determine the sparsity ratios of different layers or blocks by leveraging the global importance score, which is efficiently computed based on the zeroth-order approximation of the global model gradients. Then, the model performs local layer-wise unstructured weight pruning based on globally-informed sparsity ratios. We validate our proposed method across various multimodal and unimodal models and datasets, demonstrating significant performance improvements over prevalent pruning techniques in the high-sparsity regime.