Update app.py

app.py

@@ -32,7 +32,7 @@ llm = Llama(
     n_threads=max(1, os.cpu_count() or 2),
     n_threads_batch=max(1, os.cpu_count() or 2),
     n_batch=128,
-    n_gpu_layers=0,
+    n_gpu_layers=1,
     verbose=False,
 )