Quickstart

Installation

To install TableGPT Agent, use the following command:

pip install tablegpt-agent

This package depends on pybox to manage code execution environment. By default, pybox operates in an in-cluster mode. If you intend to run tablegpt-agent in a local environment, install the optional dependency as follows:

pip install tablegpt-agent[local]

Setup the LLM Service

Before using TableGPT Agent, ensure you have an OpenAI-compatible server configured to host TableGPT2. We recommend using vllm for this:

python -m vllm.entrypoints.openai.api_server --served-model-name TableGPT2-7B --model path/to/weights

Notes:

To analyze tabular data with tablegpt-agent, make sure TableGPT2 is served with vllm version 0.5.5 or higher.

For production environments, it's important to optimize the vllm server configuration. For details, refer to the vllm documentation on server configuration.

Chat with TableGPT Agent

To create an agent, you'll need at least an LLM instance and a PyBoxManager:

NOTE 1: This tutorial uses langchain-openai for the llm instance. Please install it first.

pip install langchain-openai

NOTE 2: TableGPT Agent fully supports aync invocation. To start a Python console that supports asynchronous operations, run the following command:

python -m asyncio

In the console, set up the agent as follows:

>>> from langchain_openai import ChatOpenAI
>>> from pybox import LocalPyBoxManager
>>> from tablegpt.agent import create_tablegpt_graph
>>> from tablegpt import DEFAULT_TABLEGPT_IPYKERNEL_PROFILE_DIR

>>> llm = ChatOpenAI(openai_api_base="YOUR_VLLM_URL", openai_api_key="whatever", model_name="TableGPT2-7B")
>>> pybox_manager = LocalPyBoxManager(profile_dir=DEFAULT_TABLEGPT_IPYKERNEL_PROFILE_DIR)

>>> agent = create_tablegpt_graph(
...     llm=llm,
...     pybox_manager=pybox_manager,
... )

To interact with the agent:

>>> from datetime import date
>>> from langchain_core.messages import HumanMessage

>>> message = HumanMessage(content="Hi")

>>> _input = {
...     "messages": [message],
...     "parent_id": "some-parent-id",
...     "date": date.today(),
... }

>>> response = await agent.ainvoke(_input)
>>> print(response["messages"])
[HumanMessage(content='Hi', additional_kwargs={}, response_metadata={}, id='7da7e51c-0ad0-4481-aa22-54acce6a82d7'), AIMessage(content="Hello! How can I assist you with your data analysis today? Do you have a dataset you'd like to work with?", additional_kwargs={'parent_id': 'some-parent-id'}, response_metadata={}, id='43df249b-9c61-44bc-a535-eb33f9efaa9e')]

You can get more detailed outputs with the astream_events method:

>>> async for event in agent.astream_events(
...     input=_input,
...     version="v2",
... ):
...     # We ignore irrelevant events here.
...     if event["event"] == "on_chat_model_end":
...         print(event["data"]["output"])
content="Hello! How can I assist you with your data analysis today? Do you have a specific dataset or problem in mind that you'd like to work on?" additional_kwargs={} response_metadata={'finish_reason': 'stop', 'model_name': 'TableGPT2-7B'} id='run-677d42b7-ae79-4695-8e54-1b02fab07427'

Full code

import asyncio
from datetime import date

from langchain_core.messages import HumanMessage
from langchain_openai import ChatOpenAI
from pybox import LocalPyBoxManager
from tablegpt.agent import create_tablegpt_graph
from tablegpt import DEFAULT_TABLEGPT_IPYKERNEL_PROFILE_DIR


# tablegpt-agent fully supports async invocation
async def main() -> None:
    llm = ChatOpenAI(openai_api_base="YOUR_VLLM_URL", openai_api_key="whatever", model_name="TableGPT2-7B")

    # Use local pybox manager for development and testing
    pybox_manager = LocalPyBoxManager(profile_dir=DEFAULT_TABLEGPT_IPYKERNEL_PROFILE_DIR)

    agent = create_tablegpt_graph(
        llm=llm,
        pybox_manager=pybox_manager,
    )

    message = HumanMessage(content="Hi")
    _input = {
        "messages": [message],
        "parent_id": "some-parent-id",
        "date": date.today(),  # noqa: DTZ011
    }

    # response = await agent.ainvoke(_input)
    # print(response["messages"])

    # More details can be obtained through the astream_events method
    async for event in agent.astream_events(
        input=_input,
        version="v2",
    ):
        print(event)  # noqa: T201


asyncio.run(main())