Update llm.py

llm.py

@@ -10,7 +10,7 @@ text_generator = pipeline(
     model="google/gemma-3n-e4b-it",
     device="cuda" if torch.cuda.is_available() else "cpu",
     torch_dtype=torch.bfloat16,
-    max_new_tokens=100  # Limit output length
+    max_new_tokens=500  # Limit output length
 )
 
 # Wrap pipeline for LangChain compatibility