TAMP: Token-Adaptive Layerwise Pruning in Multimodal Large Language Models
This work addresses the challenge of reducing model size for MLLMs, which is crucial for deployment efficiency, though it appears incremental as it builds on existing pruning methods with modality-specific adaptations.
The paper tackles the problem of pruning multimodal large language models (MLLMs) by proposing TAMP, a framework that adapts sparsity based on token diversity and input activation, achieving substantial performance improvements over existing pruning techniques on benchmarks like LLaVA-NeXT and VideoLLaMA2.
Multimodal Large Language Models (MLLMs) have shown remarkable versatility in understanding diverse multimodal data and tasks. However, these capabilities come with an increased model scale. While post-training pruning reduces model size in unimodal models, its application to MLLMs often yields limited success. Our analysis discovers that conventional methods fail to account for the unique token attributes across layers and modalities inherent to MLLMs. Inspired by this observation, we propose TAMP, a simple yet effective pruning framework tailored for MLLMs, featuring two key components: (1) Diversity-Aware Sparsity, which adjusts sparsity ratio per layer based on diversities among multimodal output tokens, preserving more parameters in high-diversity layers; and (2) Adaptive Multimodal Input Activation, which identifies representative multimodal input tokens using attention scores to guide unstructured weight pruning. We validate our method on two state-of-the-art MLLMs: LLaVA-NeXT, designed for vision-language tasks, and VideoLLaMA2, capable of processing audio, visual, and language modalities. Empirical experiments across various multimodal evaluation benchmarks demonstrate that each component of our approach substantially outperforms existing pruning techniques.