D2O: Dynamic Discriminative Operations for Efficient Long-Context Inference of Large Language Models
This addresses a critical bottleneck in deploying LLMs for long sequences, offering a practical solution for efficient inference without fine-tuning.
The paper tackles the memory inefficiency of Key-Value (KV) cache in large language models during long-context inference by introducing Dynamic Discriminative Operations (D2O), a compression method that reduces memory usage and increases inference throughput by over 3x while preserving generation quality.
Generative inference in Large Language Models (LLMs) is impeded by the growing memory demands of Key-Value (KV) cache, especially for longer sequences. Traditional KV cache eviction strategies, which discard less critical KV pairs based on attention scores, often degrade generation quality, leading to issues such as context loss or hallucinations. In this work, we introduce Dynamic Discriminative Operations (D2O), a KV cache compression method that optimizes KV cache size dynamically and discriminatively at two levels without fine-tuning, while preserving essential context. At layer level, D2O leverages the varying densities of attention weights between shallow and deep layers to dynamically determine which layers should avoid excessive eviction via a novel dynamic allocation strategy to minimize information loss. At token level, D2O incorporates a compensation mechanism that maintains a similarity threshold to re-discriminate the importance of currently discarded tokens, determining whether they should be recalled and merged with similar tokens. We conduct experiments on various benchmarks and LLM architectures. Our results show that D2O not only achieves significant memory savings and enhances inference throughput by more than 3$\times$ but also maintains high-quality long-text generation.