VL-Cache: Sparsity and Modality-Aware KV Cache Compression for Vision-Language Model Inference Acceleration

TL;DR

VL-Cache introduces a sparsity and modality-aware KV cache compression technique tailored for vision-language models, significantly reducing memory usage and accelerating inference while maintaining high accuracy.

Contribution

The paper presents a novel cache compression method for VLMs that adapts to layer-specific sparsity and token importance, outperforming existing approaches.

Findings

Retaining 10% of KV cache maintains accuracy.

Up to 2.33x end-to-end latency speedup.

GPU memory footprint reduced by 90%.

Abstract

Vision-Language Models (VLMs) have demonstrated impressive performance across a versatile set of tasks. A key challenge in accelerating VLMs is storing and accessing the large Key-Value (KV) cache that encodes long visual contexts, such as images or videos. While existing KV cache compression methods are effective for Large Language Models (LLMs), directly migrating them to VLMs yields suboptimal accuracy and speedup. To bridge the gap, we propose VL-Cache, a novel KV cache compression recipe tailored for accelerating VLM inference. In this paper, we first investigate the unique sparsity pattern of VLM attention by distinguishing visual and text tokens in prefill and decoding phases. Based on these observations, we introduce a layer-adaptive sparsity-aware cache budget allocation method that effectively distributes the limited cache budget across different layers, further reducing KV cache size without compromising accuracy. Additionally, we develop a modality-aware token scoring policy to better evaluate the token importance. Empirical results on multiple benchmark datasets demonstrate that retaining only 10% of KV cache achieves accuracy comparable to that with full cache. In a speed benchmark, our method accelerates end-to-end latency of generating 100 tokens by up to 2.33x and speeds up decoding by up to 7.08x, while reducing the memory footprint of KV cache in GPU by 90%.