Update app.py

app.py

@@ -23,7 +23,7 @@ llm = Llama(
     model_path=base_model_path,
     lora_path=adapter_path,
     n_ctx=512,       # Context length, set manually since adapter lacks it
-    n_threads=2,     # Adjust based on your system
+    n_threads=optimal_threads,     # Adjust based on your system
     use_mmap=True,
     n_gpu_layers=0    # Set to >0 if GPU acceleration is desired and supported
 )