feat: auto select llama-cpp cuda runtime - #2306

Merged

mudler merged 16 commits into

mudler:masterfrom

sozercan:auto-select-variant-cuda

May 14, 2024

Commits on May 13, 2024

auto select cpu variant
sozercan
committed
7df6910
Merge branch 'master' into auto-select-variant
sozercan
authored
ecd71b1
remove cuda target for now
sozercan
committed
e9bca66
fix metal
sozercan
committed
20b442a
fix path
sozercan
committed
c29825b
cuda
sozercan
committed
4bb8963
auto select cuda
sozercan
committed
f0a688b
update test
sozercan
committed
4ce686a
Merge branch 'master' into auto-select-variant-cuda
mudler
committed
94d96c7
select CUDA backend only if present
mudler
committed
d917ec8
ci: keep cuda bin in path
mudler
committed
4e44e00
Makefile: make dist now builds also cuda
mudler
committed
079ffeb
Keep pushing fallback in case auto-flagset/nvidia fails
mudler
committed
e7fc049
Do not build cuda on MacOS
mudler
committed
18bd7a8

Commits on May 14, 2024

cleanup
sozercan
committed
578e34f
Apply suggestions from code review
mudler
authored
6586b09