Langchain-Chatchat/configs/prompt_config.py.example

# prompt模板使用Jinja2语法，简单点就是用双大括号代替f-string的单大括号
# 本配置文件支持热加载，修改prompt模板后无需重启服务。


# LLM对话支持的变量：
#   - input: 用户输入内容

# 知识库和搜索引擎对话支持的变量：
#   - context: 从检索结果拼接的知识文本
#   - question: 用户提出的问题


PROMPT_TEMPLATES = {
    # LLM对话模板
    "llm_chat": "{{ input }}",

    # 基于本地知识问答的提示词模板
    "knowledge_base_chat":
    """
    <指令>根据已知信息，简洁和专业的来回答问题。如果无法从中得到答案，请说 “根据已知信息无法回答该问题”，不允许在答案中添加编造成分，答案请使用中文。 </指令>
    <已知信息>{{ context }}</已知信息>、
    <问题>{{ question }}</问题>""",

    # 基于agent的提示词模板
    "agent_chat":
    """
    Answer the following questions as best you can. You have access to the following tools:

    {tools}
    Use the following format:

    Question: the input question you must answer
    Thought: you should always think about what to do
    Action: the action to take, should be one of [{tool_names}]
    Action Input: the input to the action
    Observation: the result of the action
    ... (this Thought/Action/Action Input/Observation can be repeated zero or more times)
    Thought: I now know the final answer
    Final Answer: the final answer to the original input question

    Begin!

    history:
    {history}

    Question: {input}
    Thought: {agent_scratchpad}
"""
}