Update app.py

app.py

@@ -28,7 +28,7 @@ def chat(message: str, history: list, stm_state: torch.Tensor, temperature: floa
     response = ""
     for token_id in model.interact(**tokenized_query, max_seq_len=seq_len, temperature=temperature, top_p=top_p):
         response += model.stringify_token(token_id, show_memory_update=True)
-        yield history + [[message, response]]
+        yield history + [[message, response]], stm_state
     
     return history + [[message, response]], model.export_stm_state().cpu()