feat: Simplify llm setup (offload to acorn)

Author: jhwohlgemuth

.vscode/settings.json

@@ -11,12 +11,19 @@
         "statusBar.background": "#0db7ed",
         "statusBar.foreground": "#15202b",
         "statusBarItem.hoverBackground": "#0a92bd",
-        "statusBarItem.remoteBackground": "#0db7ed",
+        "statusBarItem.remoteBackground": "#d40ca3",
         "statusBarItem.remoteForeground": "#15202b",
         "titleBar.activeBackground": "#0db7ed",
         "titleBar.activeForeground": "#15202b",
         "titleBar.inactiveBackground": "#0db7ed99",
-        "titleBar.inactiveForeground": "#15202b99"
+        "titleBar.inactiveForeground": "#15202b99",
+        "activityBarTop.activeBackground": "#39c7f4",
+        "activityBarTop.background": "#39c7f4",
+        "activityBarTop.foreground": "#15202b",
+        "activityBarTop.inactiveForeground": "#15202b99",
+        "commandCenter.foreground": "#15202b",
+        "statusBar.debuggingBackground": "#0db7ed",
+        "statusBar.debuggingForeground": "#15202b"
     },
     "peacock.color": "#0db7ed",
     "files.exclude": {

config/llama-swap/config.yaml

@@ -21,21 +21,3 @@ models:
       --sleep-idle-seconds 600
       --tools all
       --ubatch-size 2048
-  "vllm-coder":
-    proxy: "http://127.0.0.1:${PORT}"
-    description: "When should the model be used?"
-    cmd: |
-      podman run
-      --name ${MODEL_ID}
-      --init
-      --rm
-      --gpus all
-      --publish ${PORT}:8000
-      -v "${env.HOME}/models:/root/models"
-      docker.io/vllm/vllm-openai:latest
-      --model ${models_dir}/Qwen2.5-0.5B-Instruct-Q4_K_M.gguf
-      --port 8000
-      --serving-name vllm-coder
-    cmdStop: podman stop ${MODEL_ID}
-    checkEndpoint: "none"
-    ttl: 0