Update app.py

app.py

@@ -205,9 +205,18 @@ def generate_response(input_dict):
         **inputs,
         max_new_tokens=512,
         temperature=0.7,
-        do_sample=True
+        do_sample=True,
+        pad_token_id=tokenizer.eos_token_id  # avoids warning
     )
-    return tokenizer.decode(outputs[0], skip_special_tokens=True).split("QUESTION:")[-1].strip()
+    decoded = tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+    # Print for debugging
+    print("🔍 Full LLM Output:\n", decoded)
+
+    
+    response_only = decoded[len(prompt):].strip()
+    return response_only
+