Hacker News
new
past
comments
ask
show
jobs
points
by
dryarzeg
13 hours ago
|
comments
by
rohansood15
5 hours ago
|
[-]
The paper is about vector quantization, which affects KV cache not model weights/sizes.
reply