Update app.py

app.py

@@ -42,7 +42,7 @@ system_prompt = f"<|im_start|>{system_role}\n{system_prompt}<|im_end|>"
 @spaces.GPU()
 def predict(message, history):
     # history = []
-    history_transformer_format = history + [[message, ""]]
+    # history_transformer_format = history + [[message, ""]]
     stop = StopOnTokens()
 
     # Formatting the input for the model.
@@ -50,9 +50,11 @@ def predict(message, history):
     #                     for item in history_transformer_format])
 
     model_messages = []
-    for item in history_transformer_format:
+    print(f'history: {history}')
+    for i, item in enumerate(history):
         model_messages.append({"role": user_role, "content": item[0]})
-        model_messages.append({"role": assistant_role, "content": item[1]})
+        if i < len(history) - 1:
+            model_messages.append({"role": assistant_role, "content": item[1]})
 
     print(f'model_messages: {model_messages}')