N
Hacker Next
login
▲
16-Bit to 1-Bit: Visual KV Cache Quantization for Efficient Multimodal LLMs
arxiv.org
87 points by
PaulHoule
54 days ago
|
1 comment
add comment
Loading comments...
kadushka 50 days ago
[-]
Have they published their code?
fdafdsfe 50 days ago
[-]
[dead]
soijaijte 50 days ago
[-]
[dead]
iwriawei 50 days ago
[-]
[dead]
fsafdsaewr 50 days ago
[-]
[flagged]