[B-7] Information Networks

11:15 AM - 11:30 AM JST(2:15 AM - 2:30 AM UTC)Candidate for Academic Incentive Award

[B-7-33] Chunked KV Cache Control for Efficient Vision-Language Model Inference

◎Kenshiro Wada¹, Kenzo Okuda¹, Hiroki Baba¹, Naoki Kimishima¹, Kentaro Hayashi¹, Tomonori Takeda¹ (1. NTT)

Keywords:

LLM,VLM,Transformer,KV Cache,Prefill,In Network Computing

Password required to view