Update app.py

app.py

@@ -17,7 +17,7 @@ def respond(message, history):
     start_time = time.time()
     outputs = model.generate(
         **inputs,
-        max_new_tokens=32,
+        max_new_tokens=16,
         do_sample=False,
         pad_token_id=tokenizer.eos_token_id
     )