提升LangChain查询分析：如何添加示例以优化LLM响应-优快云博客

本文链接：https://blog.youkuaiyun.com/python1234567_/article/details/142874035

引言

在复杂的查询分析中，语言模型（LLM）有时无法理解如何准确响应特定场景。为了提高性能，我们可以在提示中添加示例，以指导LLM更准确地生成答案。本篇文章将探讨如何为LangChain YouTube视频查询分析器添加示例，从而优化查询生成。

主要内容

安装和设置

首先，我们需要安装必要的依赖：

# 安装LangChain核心和OpenAI插件
%pip install -qU langchain-core langchain-openai

然后，设置环境变量以使用OpenAI服务：

import getpass
import os

os.environ["OPENAI_API_KEY"] = getpass.getpass()

查询模式定义

定义我们希望模型输出的查询模式，包括一个sub_queries字段，用于包含从顶层问题派生的更细化问题：

from typing import List, Optional
from langchain_core.pydantic_v1 import BaseModel, Field

class Search(BaseModel):
    query: str = Field(
        ..., description="视频转录本中应用的主相似性搜索查询。"
    )
    sub_queries: List[str] = Field(
        default_factory=list, description="分解后的相关子问题。"
    )
    publish_year: Optional[int] = Field(None, description="视频发布年份")

查询生成

构建提示模板以及LLM的结构化输出：

from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain_openai import ChatOpenAI

system = """You are an expert at converting user questions into database queries. Given a question, return a list of database queries optimized to retrieve the most relevant results."""

prompt = ChatPromptTemplate.from_messages(
    [
        ("system", system),
        MessagesPlaceholder("examples", optional=True),
        ("human", "{question}"),
    ]
)
llm = ChatOpenAI(model="gpt-3.5-turbo-0125", temperature=0)
structured_llm = llm.with_structured_output(Search)

添加示例并微调提示

通过添加示例优化查询生成结果：

examples = []

question = "What's chat langchain, is it a langchain template?"
query = Search(
    query="What is chat langchain and is it a langchain template?",
    sub_queries=["What is chat langchain", "What is a langchain template"],
)
examples.append({"input": question, "tool_calls": [query]})

将示例转化为消息的辅助函数：

import uuid
from typing import Dict
from langchain_core.messages import AIMessage, BaseMessage, HumanMessage, ToolMessage

def tool_example_to_messages(example: Dict) -> List[BaseMessage]:
    messages: List[BaseMessage] = [HumanMessage(content=example["input"])]
    openai_tool_calls = []
    for tool_call in example["tool_calls"]:
        openai_tool_calls.append(
            {
                "id": str(uuid.uuid4()),
                "type": "function",
                "function": {
                    "name": tool_call.__class__.__name__,
                    "arguments": tool_call.json(),
                },
            }
        )
    messages.append(
        AIMessage(content="", additional_kwargs={"tool_calls": openai_tool_calls})
    )
    tool_outputs = example.get("tool_outputs") or [
        "You have correctly called this tool."
    ] * len(openai_tool_calls)
    for output, tool_call in zip(tool_outputs, openai_tool_calls):
        messages.append(ToolMessage(content=output, tool_call_id=tool_call["id"]))
    return messages

example_msgs = [msg for ex in examples for msg in tool_example_to_messages(ex)]

将示例应用于查询分析器：

query_analyzer_with_examples = (
    {"question": RunnablePassthrough()}
    | prompt.partial(examples=example_msgs)
    | structured_llm
)

代码示例

以下为完整的查询调用示例：

result = query_analyzer_with_examples.invoke(
    "what's the difference between web voyager and reflection agents? do both use langgraph?"
)

print(result)