feat: wire up num_gpu config and updated clients to daemon and TUI

src/daemon/index.ts

@@ -59,10 +59,12 @@ function createModelRouter(config: Config): ModelRouter {
       localClient = new OllamaClient({
         model: models.local.model,
         host: models.local.endpoint,
+        numGpu: models.local.num_gpu,
       });
     } else if (models.local.provider === 'llamacpp') {
       localClient = new LlamaCppClient({
         endpoint: models.local.endpoint ?? 'http://localhost:8080',
+        model: models.local.model,
         authToken: models.local.auth_token,
       });
     }

src/tui.ts

@@ -48,10 +48,12 @@ function createModelRouter(config: Config): ModelRouter {
       localClient = new OllamaClient({
         model: models.local.model,
         host: models.local.endpoint,
+        numGpu: models.local.num_gpu,
       });
     } else if (models.local.provider === 'llamacpp') {
       localClient = new LlamaCppClient({
         endpoint: models.local.endpoint ?? 'http://localhost:8080',
+        model: models.local.model,
         authToken: models.local.auth_token,
       });
     }