preliminar
首先需要初始化代理并读取数据文件：from langchain_openai import ChatOpenAI
from langgraph.checkpoint.memory import MemorySaver
from pybox import LocalPyBoxManager
from tablegpt.agent import create_tablegpt_graph
llm = ChatOpenAI(openai_api_base="http://localhost:8000/v1", openai_api_key="whatever", model_name="TableGPT2-7B")
pybox_manager = LocalPyBoxManager()
checkpointer = MemorySaver()
agent = create_tablegpt_graph(llm=llm, pybox_manager=pybox_manager, checkpointer=checkpointer, session_id="my-session")
进行查询
在代理运行后，可以向其提出自然语言问题：human_message = HumanMessage(content="有多少行数据？")
response = await agent.ainvoke(
input={"messages": [human_message], "date": date.today()},
config={"configurable": {"thread_id": "my-thread"}}
)
print(response["messages"])
advertência
- 保持session_id和thread_id的一致性以确保对话连贯
- 问题应当清晰明确，最好包含具体的表格字段名
- 对于复杂问题，系统会自动分解并逐步解答
Essa resposta foi extraída do artigoTableGPT Agent: ferramenta inteligente projetada para analisar dados tabulares complexosO
Português do Brasil 
简体中文 
English 
日本語 
Deutsch