Update app.py

app.py

@@ -278,13 +278,13 @@ def generate_response():
     app.logger.info('Generator created')
 
 
-    response_tokens = []
+    response_tokens = ''
     def generate_and_log_tokens(model, generator):
         for token in generate_tokens(model, generator):
             if token == b'': # or (max_new_tokens is not None and i >= max_new_tokens):
                 log(user_request, response_tokens.decode("utf-8", errors="ignore"))
                 break
-            response_tokens.append(token)
+            response_tokens += token
             yield token
 
     # Use Response to stream tokens