update llamacpp

modules/home/programs/terminal/nvim/config/lua/lsp-config.lua

@@ -155,6 +155,9 @@ setup_lsp("lua_ls", {
 	filetypes = { "lua" },
 })
 
+-- Unison LSP Configuration
+setup_lsp("unison")
+
 -- Lua LSP Configuration
 setup_lsp("sqls", {
 	cmd = { nix_vars.sqls },

modules/nixos/services/llama-swap/config.nix

@@ -7,6 +7,33 @@ let
 in
 {
   models = {
+    # docker run --device=nvidia.com/gpu=all -v ~/.cache/huggingface:/root/.cache/huggingface -p 0.0.0.0:8081:8000 --ipc=host vllm/vllm-openai:latest --model Qwen/Qwen3-0.6B
+    # docker run --device=nvidia.com/gpu=all -v /mnt/ssd/vLLM:/root/.cache/huggingface -p 0.0.0.0:8081:8000 --ipc=host vllm/vllm-openai:latest --model cyankiwi/Devstral-Small-2-24B-Instruct-2512-AWQ-4bit --max-model-len 60000 --kv-cache-dtype fp8_e4m3
+
+    # https://huggingface.co/unsloth/GLM-4.7-Flash-GGUF/tree/main
+    "glm-4.7-flash" = {
+      name = "GLM 4.7 Flash (30B) - Thinking";
+      macros.ctx = "80000";
+      cmd = ''
+        ${llama-cpp}/bin/llama-server \
+          --port ''${PORT} \
+          -m /mnt/ssd/Models/GLM/GLM-4.7-Flash-UD-Q4_K_XL.gguf \
+          -c ''${ctx} \
+          --jinja \
+          --threads -1 \
+          --temp 0.2 \
+          --top-k 50 \
+          --top-p 0.95 \
+          --min-p 0.01 \
+          --dry-multiplier 1.1 \
+          -fit off
+      '';
+      metadata = {
+        type = [ "text-generation" ];
+      };
+      env = [ "GGML_CUDA_ENABLE_UNIFIED_MEMORY=1" ];
+    };
+
     # https://huggingface.co/unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF/tree/main
     "devstral-small-2-instruct" = {
       name = "Devstral Small 2 (24B) - Instruct";

modules/nixos/virtualisation/podman/default.nix

@@ -12,6 +12,7 @@ in
 {
   options.${namespace}.virtualisation.podman = {
     enable = lib.mkEnableOption "podman";
+    enableNvidia = lib.mkEnableOption "enable nvidia";
   };
 
   config = mkIf cfg.enable {
@@ -29,6 +30,8 @@ in
       };
     };
 
+    hardware.nvidia-container-toolkit.enable = mkIf cfg.enableNvidia true;
+
     virtualisation = {
       podman = {
         inherit (cfg) enable;

packages/llama-cpp/default.nix

@@ -7,12 +7,12 @@
   vulkanSupport = true;
 }).overrideAttrs
   (oldAttrs: rec {
-    version = "7562";
+    version = "7781";
     src = pkgs.fetchFromGitHub {
       owner = "ggml-org";
       repo = "llama.cpp";
       tag = "b${version}";
-      hash = "sha256-yuTPj41B3YitRPrD6yV25ilrIxVKebPGSqdJMpVhUDg=";
+      hash = "sha256-1ikFB4pnhSwnVPWfyQbT/7BqeyswvLNCx+aDo/ORjN0=";
       leaveDotGit = true;
       postFetch = ''
         git -C "$out" rev-parse --short HEAD > $out/COMMIT

systems/x86_64-linux/lin-va-desktop/default.nix

@@ -87,7 +87,10 @@ in
     };
 
     virtualisation = {
-      podman = enabled;
+      podman = {
+        enable = true;
+        enableNvidia = true;
+      };
     };
 
     security = {