update configuration of our ollama service

nixos/fuchsia/services/ollama/default.nix

@@ -3,6 +3,10 @@
   services.ollama = {
     enable = true;
 
+    # User and group under which to run ollama
+    user = "ollama";
+    group = "ollama";
+
     # AMD GPU Support
     acceleration = "rocm";
     # 5700xt Support
@@ -11,9 +15,29 @@
     # Language models to install
     loadModels = [
       "deepseek-r1:8b"
-      "gemma3:12b"
+      "gemma3:4b"
       "qwen3:8b"
       "llama3:8b"
+
+      # Coding models
+      "qwen2.5-coder:7b"
     ];
+
+    # Location to store models
+    models = "/srv/ollama/models";
+  };
+
+  # Enable the Open-WebUI server 
+  services.open-webui = {
+    enable = true;
+  };
+
+  # Mount our subvolume for storage of models
+  fileSystems = {
+    "/srv/ollama" = {
+      device = "/dev/disk/by-label/data";
+      fsType = "btrfs";
+      options = ["subvol=srv-ollama" "compress=zstd"];
+    };
   };
 }