Hacker News
new
past
comments
ask
show
jobs
points
by
civvv
14 hours ago
|
comments
by
7speter
14 hours ago
|
[-]
You can offload the vision projector to CPU/sysRAM
reply
by
slim
10 hours ago
|
parent
|
[-]
Awesome. How would you setup that ?
reply
by
skirmish
9 hours ago
|
parent
|
[-]
--no-mmproj-offload
See the documentation:
https://github.com/ggml-org/llama.cpp/blob/master/docs/multi...
reply