RESSA: Repair Sparse Vision-Language Models via Sparse Cross-Modality Adaptation

arXiv:2404.02424v1 Announce Type: new
Abstract: Vision-Language Models (VLMs), integrating diverse information from multiple modalities, have shown remarkable success across various tasks. However, deploying VLMs, comprising large-scale vision and language models poses challenges in resource-constrained scenarios. While pruning followed by finetuning offers a potential solution to maintain performance with smaller model sizes, its application to VLMs remains relatively unexplored, presenting two main questions: how to distribute sparsity across different modality-specific models, and how to repair the performance of pruned sparse VLMs. …

abstract arxiv challenges cs.cv cs.lg diverse finetuning however information language language models multiple performance pruning repair scale solution success tasks type via vision vision-language models vlms