Update app.py

app.py

@@ -32,7 +32,7 @@ try:
     model = AutoModelForCausalLM.from_pretrained(
         repo_name,
         device_map="auto",
-        torch_dtype=torch.float16,
+        torch_dtype=torch.bfloat16,
     )
 except Exception as e:
     print(f"Error loading model with GPU: {e}")
@@ -59,7 +59,7 @@ def generate_response(message, history):
         return "Sorry, the model could not be loaded. Please check the logs."
     
     messages = [
-        {"role": "system", "content": "You are a helpful assistant. You think loud before answering anything"},
+        {"role": "system", "content": "You are a helpful assistant. You think out loud before answering anything"},
     ]
     
     # Add chat history to messages