titaniumtown/server-config
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Compare commits

base: titaniumtown:eff5b3b8aaa4d3a889ad4de19170ed26248256b7
Branches Tags
titaniumtown:main
...
compare: titaniumtown:c83d34108eb660237feb82f1acdcf4c3d50a5400
Branches Tags
titaniumtown:main

4 Commits
eff5b3b8aa ... c83d34108e

Author SHA1 Message Date
Simon Gardling
c83d34108e Revert "llama-cpp: re-enable" 
This reverts commit e98a23934a.
 2025-10-02 22:26:30 -04:00
Simon Gardling
72d950007b llama-cpp: fix postPatch phase 2025-10-02 22:26:25 -04:00
Simon Gardling
e98a23934a llama-cpp: re-enable 2025-10-02 21:30:02 -04:00
Simon Gardling
a75f34e113 llama-cpp: change model 2025-10-02 21:29:43 -04:00
1 changed files with 9 additions and 3 deletions
Expand all files Collapse all files

12
services/llama-cpp.nix
View File

@@ -11,14 +11,20 @@
enable = true;
model = builtins.toString (
pkgs.fetchurl {
url = "https://huggingface.co/ggml-org/gpt-oss-20b-GGUF/resolve/main/gpt-oss-20b-mxfp4.gguf";
sha256 = "52f57ab7d3df3ba9173827c1c6832e73375553a846f3e32b49f1ae2daad688d4";
url = "https://huggingface.co/rodrigomt/Qwen3-30B-A3B-Thinking-Deepseek-Distill-2507-v3.1-V2-GGUF/resolve/main/Qwen3-30B-A3B-Thinking-Deepseek-Distill-2507-v3.1-V2-UD-Q4_K_XL.gguf";
sha256 = "1a3abffc8463041e24cdc43af26c99b6cfab1d2ee78fef0d793033ec0e5b58aa";
}
);
port = service_configs.ports.llama_cpp;
host = "0.0.0.0";
# vulkan broken: https://github.com/ggml-org/llama.cpp/issues/13801
package = (lib.optimizePackage inputs.llamacpp.packages.${pkgs.system}.default);
package = (
lib.optimizePackage (
inputs.llamacpp.packages.${pkgs.system}.default.overrideAttrs (old: {
postPatch = "";
})
)
);
extraFlags = [
# "-ngl"
# "9999"