GPU VM - #64

Merged

wongcallum merged 10 commits into

draft/gpu-compute-vm

Aug 27, 2026

Commits on Aug 27, 2026

gpu vm
wongcallum
committed
fb44194
use fully declarative pattern for vm-gpu
wongcallum
committed
491c0a0
set vm-gpu stateVersion to 26.05
wongcallum
committed
414f7ee
add vm-gpu to ciHosts
wongcallum
committed
5c90bf9
pin vm-gpu's standalone check to unstable nixpkgs
wongcallum
committed
2500d88
import persistence into vm-gpu
wongcallum
committed
9a66fb4
use llama-cpp service with ling-3.0-tiny
wongcallum
committed
13ca6e3
whoops
wongcallum
committed
a484371
allow llama.cpp port through firewall
wongcallum
committed
ee04a4f
llama-cpp: switch to multi-model preset, size contexts for 12GB VRAM
wongcallum
committed
17464b5