Y
Hacker News
new
|
ask
|
show
|
jobs
by
dchest
813 days ago
Usually around 5 GB for a 7B 4-bit quantized model.