MEDA: Dynamic KV Cache Allocation for Efficient Multimodal Long-Context Inference

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Wan, Zhongwei, Shen, Hui, Wang, Xin, Liu, Che, Mai, Zheda, Zhang, Mi
Format:	Preprint
Published:	2025
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2502.17599