langgraph-structured-output.ipynb

{"cells":[{"cell_type":"markdown","metadata":{"cell_id":"c4ce36f168b34553b183e4582ae32e77","deepnote_cell_type":"markdown"},"source":"# 如何使用 ReAct Agent 返回结构化输出\n\n> 参考资料：https://langchain-ai.github.io/langgraph/how-tos/react-agent-structured-output/","block_group":"83792ab1d773446196b9def57e958560"},{"cell_type":"markdown","metadata":{"cell_id":"b6826010705d4a9c8bdc64dd06c3d3ea","deepnote_cell_type":"markdown"},"source":"您可能希望您的 Agent 以结构化格式返回其输出。例如，如果 Agent 返回的输出被某些其他下游软件使用，您可能希望每次调用代理时输出都采用相同的结构化格式，以确保一致性。\n\n这里将介绍两种强制使用工具调用（Tool Calling）的 Agent 结构其输出的不同选项。我们将使用一个基本的 ReAct Agent（一个模型节点和一个工具调用节点）以及一个在最后的节点，该节点将格式化用户响应。这两种选项将使用与下面图表相同的图结构，但在内部机制上有所不同。\n\n![](https://s2.loli.net/2024/09/19/GMUi9ZJAjupv6Cd.png)","block_group":"5f75ce4387b04e2b93d88ad7d00a94ea"},{"cell_type":"markdown","metadata":{"cell_id":"cf6a48f28dcc415aa0bd2b16209b63be","deepnote_cell_type":"markdown"},"source":"## 选项 1：将输出绑定为工具\n\n![](https://s2.loli.net/2024/09/19/cIHJgtGD2nz9asi.png)\n\n您可以通过将您希望使用的输出作为附加工具绑定到 `agent` 节点来强制您的工具调用代理具有结构化输出。\n\n与基本 ReAct 代理不同，在这种情况下，`agent` 节点不是在 `tools` 和 `END` 之间进行选择，而是选择它调用的特定工具。在这种情况下，预期的流程是 `agent` 节点中的 LLM 首先选择动作工具，在接收到动作工具输出后，将调用响应工具，然后将其路由到仅对 `agent` 节点工具调用中的参数进行结构化的 `respond` 节点。\n\n### 优点和缺点\n\n这种格式的优点是只需要一个 LLM，因此可以节省金钱和延迟。\n\n这种选择的缺点是并不能保证当需要时，单个 LLM 会调用正确的工具。当我们使用 `bind_tools` 时，我们可以通过将 `tool_choice` 设置为 `any` 来帮助 LLM，这会强制 LLM 在每一轮至少选择一个工具，但这远非万无一失的策略。此外，另一个缺点是 Agent 可能会调用多个工具，因此我们需要在我们的路由函数中明确检查这一点（或者如果我们使用 OpenAI，我们可以设置 `parallell_tool_calling=False` 以确保一次只调用一个工具）。","block_group":"ecb26a96a5a841c8b8b611cc5c69d192"},{"cell_type":"markdown","metadata":{"formattedRanges":[],"cell_id":"4f74762c549c47cfad4b2c084df6bbe7","deepnote_cell_type":"text-cell-callout"},"source":"> 我们将在本教程中使用一个工具来查找天气，并将返回一个结构化的天气响应给用户。","block_group":"ecb2fe47a90f482c93fb6baa9990f0a6"},{"cell_type":"code","metadata":{"source_hash":"c099352e","execution_start":1730871334469,"execution_millis":0,"execution_context_id":"7b6a3e18-120c-436c-96ec-7d7316f81f10","cell_id":"5630f7198ad74223a82a69b430ae1e31","deepnote_cell_type":"code"},"source":"from pydantic import BaseModel, Field\nfrom typing import Literal\nfrom langchain_core.tools import tool\nfrom langchain_openai import ChatOpenAI\nfrom langgraph.graph import MessagesState\n\nclass WeatherResponse(BaseModel):\n    \"\"\"Respond to the user with this\"\"\"\n    temperature: float = Field(description=\"The temperature in fahrenheit\")\n    wind_directon: str = Field(description=\"The direction of the wind in abbreviated form\")\n    wind_speed: float = Field(description=\"The speed of the wind in km/h\")\n\n# Inherit 'messages' key from MessagesState, which is a list of chat messages \nclass AgentState(MessagesState):\n    # Final structured response from the agent\n    final_response: WeatherResponse\n\n@tool\ndef get_weather(city: Literal[\"nyc\", \"sf\"]):\n    \"\"\"Use this to get weather information.\"\"\"\n    if city == \"nyc\":\n        return \"It is cloudy in NYC, with 5 mph winds in the North-East direction and a temperature of 70 degrees\"\n    elif city == \"sf\":\n        return \"It is 75 degrees and sunny in SF, with 3 mph winds in the South-East direction\"\n    else:\n        raise AssertionError(\"Unknown city\")\n    \nmodel = ChatOpenAI(model=\"gpt-4o-mini\")\n    \nmodel_with_tools = model.bind_tools([get_weather])\nmodel_with_structured_output = model.with_structured_output(WeatherResponse)","block_group":"5630f7198ad74223a82a69b430ae1e31","execution_count":11,"outputs":[],"outputs_reference":null,"content_dependencies":null},{"cell_type":"code","metadata":{"source_hash":"d6896da8","execution_start":1730870657872,"execution_millis":250,"execution_context_id":"7b6a3e18-120c-436c-96ec-7d7316f81f10","cell_id":"865155decfd2496a9b8f4c8f519a7ac1","deepnote_cell_type":"code"},"source":"from langgraph.graph import StateGraph, END\nfrom langgraph.prebuilt import ToolNode\n\n# Force the model to use tools by passing tool_choice=\"any\"    \nmodel_with_response_tool = model.bind_tools([get_weather, WeatherResponse],tool_choice=\"any\")\n\n# Define the function that calls the model\ndef call_model(state: AgentState):\n    response = model_with_response_tool.invoke(state['messages'])\n    # We return a list, because this will get added to the existing list\n    return {\"messages\": [response]}\n\n# Define the function that responds to the user\ndef respond(state: AgentState):\n    # Construct the final answer from the arguments of the last tool call\n    response = WeatherResponse(**state['messages'][-1].tool_calls[0]['args'])\n    # We return the final answer\n    return {\"final_response\": response}\n\n# Define the function that determines whether to continue or not\ndef should_continue(state: AgentState):\n    messages = state[\"messages\"]\n    last_message = messages[-1]\n    # If there is only one tool call and it is the response tool call we respond to the user\n    if len(last_message.tool_calls) == 1 and last_message.tool_calls[0]['name'] == \"WeatherResponse\":\n        return \"respond\"\n    # Otherwise we will use the tool node again\n    else:\n        return \"continue\"\n\n# Define a new graph\nworkflow = StateGraph(AgentState)\n\n# Define the two nodes we will cycle between\nworkflow.add_node(\"agent\", call_model)\nworkflow.add_node(\"respond\", respond)\nworkflow.add_node(\"tools\", ToolNode(tools))\n\n# Set the entrypoint as `agent`\n# This means that this node is the first one called\nworkflow.set_entry_point(\"agent\")\n\n# We now add a conditional edge\nworkflow.add_conditional_edges(\n    \"agent\",\n    should_continue,\n    {\n        \"continue\": \"tools\",\n        \"respond\": \"respond\",\n    },\n)\n\nworkflow.add_edge(\"tools\", \"agent\")\nworkflow.add_edge(\"respond\", END)\ngraph = workflow.compile()","block_group":"f77465a5db3240b799808149d5431a98","execution_count":2,"outputs":[],"outputs_reference":null,"content_dependencies":null},{"cell_type":"code","metadata":{"source_hash":"e501684d","execution_start":1730870665646,"execution_millis":544,"execution_context_id":"7b6a3e18-120c-436c-96ec-7d7316f81f10","cell_id":"d2e8c331655d491e8b11a90ebebdbbdc","deepnote_cell_type":"code"},"source":"from IPython.display import Image\n\nImage(graph.get_graph().draw_mermaid_png())","block_group":"d788aaa45b594c6092a1e5785fbcba15","execution_count":3,"outputs":[{"output_type":"execute_result","execution_count":3,"data":{"image/jpeg":"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","text/plain":"<IPython.core.display.Image object>"},"metadata":{}}],"outputs_reference":"s3:deepnote-cell-outputs-production/f3bffafb-2c50-437c-9147-50d0cbc252f6","content_dependencies":null},{"cell_type":"code","metadata":{"source_hash":"28d7a243","execution_start":1730870672262,"execution_millis":3561,"execution_context_id":"7b6a3e18-120c-436c-96ec-7d7316f81f10","cell_id":"19e3000665e2476c9d1e852105893cdb","deepnote_cell_type":"code"},"source":"graph.invoke(input={\"messages\": [(\"human\", \"what's the weather in SF?\")]})['final_response']","block_group":"b7b4905fcb5c45aab713a88c29d6eda3","execution_count":4,"outputs":[{"output_type":"execute_result","execution_count":4,"data":{"text/plain":"WeatherResponse(temperature=75.0, wind_directon='SE', wind_speed=3.0)"},"metadata":{}}],"outputs_reference":"dbtable:cell_outputs/fbbd07fc-e776-4dcd-81ae-97b81df54d71","content_dependencies":null},{"cell_type":"markdown","metadata":{"formattedRanges":[{"url":"https://smith.langchain.com/public/eecbc1f6-a063-492f-9384-5e20597cf2a3/r","type":"link","ranges":[],"toCodePoint":92,"fromCodePoint":19}],"cell_id":"7438b12ffe8f4c61ac0816afa92ae010","deepnote_cell_type":"text-cell-p"},"source":"👆 LangSmith Trace: https://smith.langchain.com/public/eecbc1f6-a063-492f-9384-5e20597cf2a3/r","block_group":"8d473ccf5414476292f385e3f4c86ff0"},{"cell_type":"markdown","metadata":{"cell_id":"66493845debc403ebf7af67b655a6cd3","deepnote_cell_type":"markdown"},"source":"## 选项 2：使用第二个 LLM 来强制输出\n\n![](https://s2.loli.net/2024/09/19/UfEnFlDmZ9VzdBP.png)\n\n在这种情况下，您将像平常一样定义一个基本的 ReAct Agent，但不是让 `agent` 节点在 `tools` 节点和结束对话之间选择，而是让 `agent` 节点在 `tools` 节点和 `respond` 节点之间选择。 `respond` 节点将包含一个使用结构化输出的第二个 LLM，一旦被调用，将直接返回给用户。\n\n您可以将这种方法视为在响应用户之前的一个额外步骤的基本 ReAct。\n\n### 优点和缺点\n\n这种方法的好处是它保证了结构化的输出（只要 `.with_structured_output` 与 LLM 按预期工作）。\n\n使用这种方法的缺点是需要在响应用户之前进行额外的 LLM 调用，这可能会增加成本以及延迟。此外，由于没有向 `agent` 节点 LLM 提供关于所需输出模式的信息，存在风险，即 `agent` LLM 将无法调用回答正确输出模式所需的正确工具。","block_group":"6feb37be1a2945e59265357046d29a5d"},{"cell_type":"code","metadata":{"source_hash":"bea98506","execution_start":1730871344797,"execution_millis":0,"execution_context_id":"7b6a3e18-120c-436c-96ec-7d7316f81f10","cell_id":"78fd2bfaec7c4625b38b715f9dcd28ca","deepnote_cell_type":"code"},"source":"from langgraph.graph import StateGraph, END\nfrom langgraph.prebuilt import ToolNode\nfrom langchain_core.messages import HumanMessage\n\n# Define the function that calls the model\ndef call_model(state: AgentState):\n    response = model_with_tools.invoke(state['messages'])\n    # We return a list, because this will get added to the existing list\n    return {\"messages\": [response]}\n\n# Define the function that responds to the user\ndef respond(state: AgentState):\n    # We call the model with structured output in order to return the same format to the user every time\n    # state['messages'][-2] is the last ToolMessage in the convo, which we convert to a HumanMessage for the model to use\n    # We could also pass the entire chat history, but this saves tokens since all we care to structure is the output of the tool\n    response = model_with_structured_output.invoke([HumanMessage(content=state['messages'][-2].content)])\n    # We return the final answer\n    return {\"final_response\": response}\n\n# Define the function that determines whether to continue or not\ndef should_continue(state: AgentState):\n    messages = state[\"messages\"]\n    last_message = messages[-1]\n    # If there is no function call, then we respond to the user\n    if not last_message.tool_calls:\n        return \"respond\"\n    # Otherwise if there is, we continue\n    else:\n        return \"continue\"\n\n# Define a new graph\nworkflow = StateGraph(AgentState)\n\n# Define the two nodes we will cycle between\nworkflow.add_node(\"agent\", call_model)\nworkflow.add_node(\"respond\", respond)\nworkflow.add_node(\"tools\", ToolNode(tools))\n\n# Set the entrypoint as `agent`\n# This means that this node is the first one called\nworkflow.set_entry_point(\"agent\")\n\n# We now add a conditional edge\nworkflow.add_conditional_edges(\n    \"agent\",\n    should_continue,\n    {\n        \"continue\": \"tools\",\n        \"respond\": \"respond\",\n    },\n)\n\nworkflow.add_edge(\"tools\", \"agent\")\nworkflow.add_edge(\"respond\", END)\ngraph = workflow.compile()","block_group":"dc058906cc704248865db8c24b8f48a5","execution_count":12,"outputs":[],"outputs_reference":null,"content_dependencies":null},{"cell_type":"code","metadata":{"source_hash":"e501684d","execution_start":1730870840072,"execution_millis":29,"execution_context_id":"7b6a3e18-120c-436c-96ec-7d7316f81f10","cell_id":"6899e1a0de5246a6891f48158767da2a","deepnote_cell_type":"code"},"source":"from IPython.display import Image\n\nImage(graph.get_graph().draw_mermaid_png())","block_group":"0f627a589ce947c092aab2d0662f470c","execution_count":7,"outputs":[{"output_type":"execute_result","execution_count":7,"data":{"image/jpeg":"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","text/plain":"<IPython.core.display.Image object>"},"metadata":{}}],"outputs_reference":"s3:deepnote-cell-outputs-production/ca5e9f68-35a2-4613-9543-994f9a709a73","content_dependencies":null},{"cell_type":"code","metadata":{"source_hash":"28d7a243","execution_start":1730871350077,"execution_millis":4683,"execution_context_id":"7b6a3e18-120c-436c-96ec-7d7316f81f10","cell_id":"e107ffe0a1a1402c80ff9e58a0ba4e90","deepnote_cell_type":"code"},"source":"graph.invoke(input={\"messages\": [(\"human\", \"what's the weather in SF?\")]})['final_response']","block_group":"dae6c688479540588f9cd1de362755aa","execution_count":13,"outputs":[{"output_type":"execute_result","execution_count":13,"data":{"text/plain":"WeatherResponse(temperature=75.0, wind_directon='SE', wind_speed=3.0)"},"metadata":{}}],"outputs_reference":"dbtable:cell_outputs/f2808083-ec5f-4c9d-8242-9be6cfdd5776","content_dependencies":null},{"cell_type":"markdown","metadata":{"formattedRanges":[{"url":"https://smith.langchain.com/public/ee20b5dd-c7b5-4832-a271-15fd1642a21d/r","type":"link","ranges":[],"toCodePoint":92,"fromCodePoint":19}],"cell_id":"5ecf812483404d04bace49f5875b638c","deepnote_cell_type":"text-cell-p"},"source":"👆 LangSmith Trace: https://smith.langchain.com/public/ee20b5dd-c7b5-4832-a271-15fd1642a21d/r","block_group":"02f558be44df4ab7b7c5bd40cd89a2b6"},{"cell_type":"markdown","source":"<a style='text-decoration:none;line-height:16px;display:flex;color:#5B5B62;padding:10px;justify-content:end;' href='https://deepnote.com?utm_source=created-in-deepnote-cell&projectId=e6124301-0a3c-4c43-a70b-884a597351fa' target=\"_blank\">\n<img alt='Created in deepnote.com' style='display:inline;max-height:16px;margin:0px;margin-right:7.5px;' src='data:image/svg+xml;base64,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' > </img>\nCreated in <span style='font-weight:600;margin-left:4px;'>Deepnote</span></a>","metadata":{"created_in_deepnote_cell":true,"deepnote_cell_type":"markdown"}}],"nbformat":4,"nbformat_minor":0,"metadata":{"deepnote_notebook_id":"a7e6925c8b8d4e769c7b4f232d22bf96"}}