Update app.py

app.py

@@ -866,7 +866,7 @@ Consider the conversation context to provide a helpful response that acknowledge
                 do_sample=True,
                 eos_token_id=tokenizer.eos_token_id,
                 pad_token_id=tokenizer.eos_token_id,
-                repetition_penalty=1.1
+                repetition_penalty=1.1,
                 attention_mask=None,    # ✅ Let model handle this
                 use_cache=True         # ✅ Add this for efficiency