ReFreeKV: Towards Threshold-Free KV Cache Compression

ReFreeKV: Towards Threshold-Free KV Cache Compression

ReFreeKV addresses the limitations of threshold-dependent KV cache pruning by introducing a threshold-free approach that adaptively allocates compression budgets while maintaining…

Hugging Face · Daily Papers ·Xuanfan Ni, Liyan Xu · ·▲ 43 upvotes

Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.

Autores: Xuanfan Ni, Liyan Xu, Chenyang Lyu, Longyue Wang, Mo Yu, Lemao Liu

  • 43 upvotes da comunidade
  • Temas: KV cache pruning, threshold-free methods, adaptive budget allocation, KV cache compression, LLM inference

Resumo

Resumo original (em inglês), extraído do paper:

ReFreeKV addresses the limitations of threshold-dependent KV cache pruning by introducing a threshold-free approach that adaptively allocates compression budgets while maintaining full-cache performance across diverse datasets and model sizes.

Onde ler

compartilhar: