我agent中在模仿SteamshipImageGenerationTool写自带参数的tool,结果异常,但方法功能直接在chain中调用是正常运行的

[guyouyue]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

async def agent_chat(chat_history_list_chunk, model_name, message_list,
                     session_id, task_id):
    llm_session, model_id = llm_utils.get_easy_openai_llm_session(model_name)
    tools = [
        # SteamshipImageGenerationTool(model_name="dall-e"),
        RoleDirectChatTool(chat_history_list_chunk=chat_history_list_chunk, model_name=model_name,
                           message_list=message_list, session_id=session_id, task_id=task_id),
    ]
    agent_executor = initialize_agent(agent=AgentType.STRUCTURED_CHAT_ZERO_SHOT_REACT_DESCRIPTION, llm=llm_session, tools=tools,
                                      verbose=True,
                                      callbacks=[
                                          SqlStdOutCallbackHandler(session_id, task_id, model_id),
                                      ],
                                      return_intermediate_steps=True, )
    bot_content = agent_executor.run(
        message_list,
        session_id,
        task_id,
        chat_history_list_chunk,
        model_name,
        model_id,
        version="v1"
    )
    return bot_content.get("output")
class RoleDirectChatTool(BaseTool):
    name = RoleDirectChatTool_NAME
    description = RoleDirectChatTool_description
    chat_history_list_chunk: str
    model_name: str
    message_list: list
    session_id: int
    task_id: str

    # def _run(self, chat_history_list_chunk, model_name, message_list, session_id, task_id) -> [ContentMessage]:
    #     return cuser_content_direct_chat(chat_history_list_chunk, model_name, message_list,
    #                                      session_id, task_id)
    def _run(self) -> [ContentMessage]:
        return cuser_content_direct_chat(self.chat_history_list_chunk, self.model_name, self.message_list,
                                         self.session_id, self.task_id)

Description

我agent中在模仿SteamshipImageGenerationTool写自带参数的tool,结果异常,但方法功能直接在chain中调用是正常运行的

System Info

python -m langchain_core.sys_info

[dosubot[bot]]

@guyouyue Hello there! I'm here to help you with any bugs, questions, or contributions you have. Let's work together to resolve the issue you're facing.

要在你的 agent 中模仿 SteamshipImageGenerationTool 创建自带参数的工具，并避免异常，可以参考以下步骤：

1. 确保工具类的定义正确：确保你的 RoleDirectChatTool 类正确继承了 BaseTool 并且所有必要的参数都已定义。

2. 验证参数：在工具类中使用 root_validator 来验证参数是否正确。

3. 初始化工具：在 initialize_agent 函数中正确初始化工具，并确保所有依赖项都已正确导入和配置。

以下是一个示例，展示了如何定义和使用自定义工具：

from langchain_core.tools import BaseTool
from langchain_core.pydantic_v1 import root_validator
from typing import List

class RoleDirectChatTool(BaseTool):
    name: str = "RoleDirectChatTool"
    description: str = "A tool for direct chat with role-based context."
    chat_history_list_chunk: str
    model_name: str
    message_list: List[str]
    session_id: int
    task_id: str

    @root_validator(pre=True)
    def validate_params(cls, values):
        if not values.get("chat_history_list_chunk"):
            raise ValueError("chat_history_list_chunk is required")
        if not values.get("model_name"):
            raise ValueError("model_name is required")
        if not values.get("message_list"):
            raise ValueError("message_list is required")
        if not values.get("session_id"):
            raise ValueError("session_id is required")
        if not values.get("task_id"):
            raise ValueError("task_id is required")
        return values

    def _run(self) -> List[str]:
        return cuser_content_direct_chat(
            self.chat_history_list_chunk, 
            self.model_name, 
            self.message_list,
            self.session_id, 
            self.task_id
        )

# 初始化 agent
async def agent_chat(chat_history_list_chunk, model_name, message_list, session_id, task_id):
    llm_session, model_id = llm_utils.get_easy_openai_llm_session(model_name)
    tools = [
        RoleDirectChatTool(
            chat_history_list_chunk=chat_history_list_chunk, 
            model_name=model_name,
            message_list=message_list, 
            session_id=session_id, 
            task_id=task_id
        ),
    ]
    agent_executor = initialize_agent(
        agent=AgentType.STRUCTURED_CHAT_ZERO_SHOT_REACT_DESCRIPTION, 
        llm=llm_session, 
        tools=tools,
        verbose=True,
        callbacks=[SqlStdOutCallbackHandler(session_id, task_id, model_id)],
        return_intermediate_steps=True
    )
    bot_content = agent_executor.run(
        message_list,
        session_id,
        task_id,
        chat_history_list_chunk,
        model_name,
        model_id,
        version="v1"
    )
    return bot_content.get("output")

确保在使用 RoleDirectChatTool 时，所有必需的参数都已正确传递，并且工具类中的 _run 方法能够正确调用 cuser_content_direct_chat 函数。如果仍然遇到异常，请检查具体的错误信息和堆栈跟踪，以便进一步调试。

To continue talking to Dosu, mention @dosu.