mirror of
https://github.com/ikawrakow/ik_llama.cpp.git
synced 2026-02-10 08:20:09 +00:00
0ceeb11721
* Merging mainline - WIP * Merging mainline - WIP AVX2 and CUDA appear to work. CUDA performance seems slightly (~1-2%) lower as it is so often the case with llama.cpp/ggml after some "improvements" have been made. * Merging mainline - fix Metal * Remove check --------- Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
22 lines
434 B
Nix
22 lines
434 B
Nix
{
|
|
perSystem =
|
|
{ config, lib, ... }:
|
|
{
|
|
apps =
|
|
let
|
|
inherit (config.packages) default;
|
|
binaries = [
|
|
"llama-cli"
|
|
"llama-embedding"
|
|
"llama-server"
|
|
"llama-quantize"
|
|
];
|
|
mkApp = name: {
|
|
type = "app";
|
|
program = "${default}/bin/${name}";
|
|
};
|
|
in
|
|
lib.genAttrs binaries mkApp;
|
|
};
|
|
}
|