### What problem does this PR solve? ### Type of change - [x] New Feature (non-breaking change which adds functionality)tags/v0.13.0
| @@ -189,6 +189,7 @@ def chat(dialog, messages, stream=True, **kwargs): | |||
| used_token_count, msg = message_fit_in(msg, int(max_tokens * 0.97)) | |||
| assert len(msg) >= 2, f"message_fit_in has bug: {msg}" | |||
| prompt = msg[0]["content"] | |||
| prompt += "\n\n### Query:\n%s" % " ".join(questions) | |||
| if "max_tokens" in gen_conf: | |||
| gen_conf["max_tokens"] = min( | |||