Code Interpreter APIなるパッケージを触ってみる

背景

Code InterpreterはChatGPT上でファイルに対して処理が可能な機能としてリリースされました。
ただし、他の記事にもあるように、サーバーの制約が多いことから、ローカルで同機能が実施できるかを探していたところ、LangChainを活用したパッケージを公開している人がいたので、実施してみました。

結論

　実行自体はできず、LangChainでの制約？なのか「GPT-4」が利用できませんでした。
　どなたか分かる方、ご教示ください！

概要

このJupyter Notebookは、codeinterpreter-api というAPIの使用方法を示すものです。
参考として、公式のblog も提供されています。

内容の要点：

1. os ライブラリを使用して、OpenAIのAPIキーを環境変数として設定。
2. CodeInterpreterSession を使用して、「テクジャイアンツ（Apple、Google、Microsoft、Amazon）の2023年の相対的なパフォーマンスをプロットする」ための指示をAIに送信。
3. AIからのレスポンスを表示し、関連するイメージファイルも表示。

環境

　Colab(2023.08.10 時点)

実行プログラム

!pip install "codeinterpreterapi[all]"

import os

os.environ["OPENAI_API_KEY"] = "sk-XXXc"
os.environ["VERBOSE"] = "True"

from codeinterpreterapi import CodeInterpreterSession

async with CodeInterpreterSession() as session:
    response = await session.generate_response(
    """
        Plot the relative performance of tech giants
        (Apple, Google, Microsoft, Amazon) in 2023.
        Use a normalized plot so we can easily compare
        the percentage changes of each stock's price over the period.
    """
    )

    print("AI: ", response.content)
    for file in response.files:
        file.show_image()

INFO: Using a LocalBox which is not fully isolated
      and not scalable across multiple users.
      Make sure to use a CODEBOX_API_KEY in production.
      Set envar SHOW_INFO=False to not see this again.

Starting kernel...
Waiting for kernel to start...
DEPRECATION WARNING: Use agenerate_response for async generation.
This function will be converted to sync in the future.
You can use generate_response_sync for now.


[1m> Entering new AgentExecutor chain...[0m
AI:  Sorry, something went while generating your response.Please try again or restart the session.


Traceback (most recent call last):
  File "/usr/local/lib/python3.10/dist-packages/codeinterpreterapi/session.py", line 437, in agenerate_response
    response = await self.agent_executor.arun(input=user_request.content)
  File "/usr/local/lib/python3.10/dist-packages/langchain/chains/base.py", line 517, in arun
    await self.acall(
  File "/usr/local/lib/python3.10/dist-packages/langchain/chains/base.py", line 310, in acall
    raise e
  File "/usr/local/lib/python3.10/dist-packages/langchain/chains/base.py", line 304, in acall
    await self._acall(inputs, run_manager=run_manager)
  File "/usr/local/lib/python3.10/dist-packages/langchain/agents/agent.py", line 1078, in _acall
    next_step_output = await self._atake_next_step(
  File "/usr/local/lib/python3.10/dist-packages/langchain/agents/agent.py", line 925, in _atake_next_step
    output = await self.agent.aplan(
  File "/usr/local/lib/python3.10/dist-packages/codeinterpreterapi/agents/functions_agent.py", line 253, in aplan
    predicted_message = await self.llm.apredict_messages(
  File "/usr/local/lib/python3.10/dist-packages/langchain/chat_models/base.py", line 429, in apredict_messages
    return await self._call_async(messages, stop=_stop, **kwargs)
  File "/usr/local/lib/python3.10/dist-packages/langchain/chat_models/base.py", line 369, in _call_async
    result = await self.agenerate(
  File "/usr/local/lib/python3.10/dist-packages/langchain/chat_models/base.py", line 205, in agenerate
    raise exceptions[0]
  File "/usr/local/lib/python3.10/dist-packages/langchain/chat_models/base.py", line 306, in _agenerate_with_cache
    return await self._agenerate(
  File "/usr/local/lib/python3.10/dist-packages/langchain/chat_models/openai.py", line 416, in _agenerate
    response = await acompletion_with_retry(
  File "/usr/local/lib/python3.10/dist-packages/langchain/chat_models/openai.py", line 95, in acompletion_with_retry
    return await _completion_with_retry(**kwargs)
  File "/usr/local/lib/python3.10/dist-packages/tenacity/_asyncio.py", line 88, in async_wrapped
    return await fn(*args, **kwargs)
  File "/usr/local/lib/python3.10/dist-packages/tenacity/_asyncio.py", line 47, in __call__
    do = self.iter(retry_state=retry_state)
  File "/usr/local/lib/python3.10/dist-packages/tenacity/__init__.py", line 314, in iter
    return fut.result()
  File "/usr/lib/python3.10/concurrent/futures/_base.py", line 451, in result
    return self.__get_result()
  File "/usr/lib/python3.10/concurrent/futures/_base.py", line 403, in __get_result
    raise self._exception
  File "/usr/local/lib/python3.10/dist-packages/tenacity/_asyncio.py", line 50, in __call__
    result = await fn(*args, **kwargs)
  File "/usr/local/lib/python3.10/dist-packages/langchain/chat_models/openai.py", line 93, in _completion_with_retry
    return await llm.client.acreate(**kwargs)
  File "/usr/local/lib/python3.10/dist-packages/openai/api_resources/chat_completion.py", line 45, in acreate
    return await super().acreate(*args, **kwargs)
  File "/usr/local/lib/python3.10/dist-packages/openai/api_resources/abstract/engine_api_resource.py", line 217, in acreate
    response, _, api_key = await requestor.arequest(
  File "/usr/local/lib/python3.10/dist-packages/openai/api_requestor.py", line 382, in arequest
    resp, got_stream = await self._interpret_async_response(result, stream)
  File "/usr/local/lib/python3.10/dist-packages/openai/api_requestor.py", line 726, in _interpret_async_response
    self._interpret_response_line(
  File "/usr/local/lib/python3.10/dist-packages/openai/api_requestor.py", line 763, in _interpret_response_line
    raise self.handle_error_response(
openai.error.InvalidRequestError: The model `gpt-4` does not exist or you do not have access to it. Learn more: https://help.openai.com/en/articles/7102672-how-can-i-access-gpt-4.

備考

　パッケージのサンプルコードを見るとhttp URLを指定しプロンプト上に記載すると、データの取得が可能なようなので、ChatGPT上よりかは汎化できる気がします。