Update app.py

app.py

@@ -17,7 +17,7 @@ llm = Llama(
     model_path=base_model_path,
     lora_path=adapter_path,
     n_ctx=256,          # Reduced context length (adjust based on your needs)
-    n_threads=8,        # Use all available CPU cores (adjust based on your system)
+    n_threads=2,        # Use all available CPU cores (adjust based on your system)
     n_gpu_layers=0,    # Enable GPU acceleration (adjust based on VRAM)
     use_mmap=False,     # Disable memory mapping for faster access
     use_mlock=True,     # Lock model in memory for stability