langchain - 💡(How to fix) Fix LangChain Agent + vLLM Qwen2-VL + @tool [4 comments, 2 participants]

# from dataclasses import dataclass

# from langchain.agents import create_agent
# from langchain.chat_models import init_chat_model
# from langchain.tools import tool, ToolRuntime
# from langgraph.checkpoint.memory import InMemorySaver
# from langchain.agents.structured_output import ToolStrategy
# from langchain_openai import ChatOpenAI

# # Define system prompt
# SYSTEM_PROMPT = """You are an expert weather forecaster, who speaks in puns.

# You have access to two tools:

# - get_weather_for_location: use this to get the weather for a specific location
# - get_user_location: use this to get the user's location

# If a user asks you for the weather, make sure you know the location. If you can tell from the question that they mean wherever they are, use the get_user_location tool to find their location."""

# # Define context schema
# @dataclass
# class Context:
#     """Custom runtime context schema."""
#     user_id: str

# # Define tools
# @tool
# def get_weather_for_location(city: str) -> str:
#     """Get weather for a given city."""
#     return f"It's always sunny in {city}!"

# @tool
# def get_user_location(runtime: ToolRuntime[Context]) -> str:
#     """Retrieve user information based on user ID."""
#     user_id = runtime.context.user_id
#     return "Florida" if user_id == "1" else "SF"

# # Configure model
# model = ChatOpenAI(
#     base_url="http://localhost:40050/v1",
#     model="Qwen/Qwen2-VL-7B-Instruct",
#     temperature=0.1,
#     api_key="EMPTY"
# )

# # Define response format
# @dataclass
# class ResponseFormat:
#     """Response schema for the agent."""
#     # A punny response (always required)
#     punny_response: str
#     # Any interesting information about the weather if available
#     weather_conditions: str | None = None

# # Set up memory
# checkpointer = InMemorySaver()

# # Create agent
# agent = create_agent(
#     model=model,
#     system_prompt=SYSTEM_PROMPT,
#     tools=[get_user_location, get_weather_for_location],
#     context_schema=Context,
#     response_format=ToolStrategy(ResponseFormat),
#     checkpointer=checkpointer
# )

# # Run agent
# # `thread_id` is a unique identifier for a given conversation.
# config = {"configurable": {"thread_id": "1"}}

# response = agent.invoke(
#     {"messages": [{"role": "user", "content": "what is the weather outside?"}]},
#     config=config,
#     context=Context(user_id="1")
# )

# print(response['structured_response'])
# # ResponseFormat(
# #     punny_response="Florida is still having a 'sun-derful' day! The sunshine is playing 'ray-dio' hits all day long! I'd say it's the perfect weather for some 'solar-bration'! If you were hoping for rain, I'm afraid that idea is all 'washed up' - the forecast remains 'clear-ly' brilliant!",
# #     weather_conditions="It's always sunny in Florida!"
# # )


# # Note that we can continue the conversation using the same `thread_id`.
# response = agent.invoke(
#     {"messages": [{"role": "user", "content": "thank you!"}]},
#     config=config,
#     context=Context(user_id="1")
# )

# print(response['structured_response'])
# # ResponseFormat(
# #     punny_response="You're 'thund-erfully' welcome! It's always a 'breeze' to help you stay 'current' with the weather. I'm just 'cloud'-ing around waiting to 'shower' you with more forecasts whenever you need them. Have a 'sun-sational' day in the Florida sunshine!",
# #     weather_conditions=None
# # )

# from dataclasses import dataclass
# from langchain_openai import ChatOpenAI
# from langgraph.checkpoint.memory import InMemorySaver
# from langchain.agents.structured_output import ToolStrategy
# from langchain.agents import create_agent
# import time

# # 系统提示
# SYSTEM_PROMPT = """
# You are a weather forecaster who speaks in puns.
# Answer in a concise, structured way.
# """

# # 上下文
# @dataclass
# class Context:
#     user_id: str

# # 模型
# model = ChatOpenAI(
#     base_url="http://localhost:40050/v1",
#     model="Qwen/Qwen2-VL-7B-Instruct",
#     temperature=0.1,
#     api_key="EMPTY"
# )

# # 输出格式
# @dataclass
# class ResponseFormat:
#     punny_response: str
#     weather_conditions: str | None = None

# # 内存
# checkpointer = InMemorySaver()

# # 创建 Agent
# agent = create_agent(
#     model=model,
#     system_prompt=SYSTEM_PROMPT,
#     tools=[],  # 暂时不使用工具
#     context_schema=Context,
#     response_format=ToolStrategy(ResponseFormat),
#     checkpointer=checkpointer
# )

# # 线程配置
# config = {"configurable": {"thread_id": "1"}}

# # ------------------------
# # 辅助函数：将结构化输出包装为 OpenAI 风格
# # ------------------------
# def wrap_openai_format(structured_response: ResponseFormat):
#     return {
#         "id": f"chatcmpl-{int(time.time())}",
#         "object": "chat.completion",
#         "created": int(time.time()),
#         "model": "Qwen/Qwen2-VL-7B-Instruct",
#         "choices": [
#             {
#                 "index": 0,
#                 "message": {
#                     "role": "assistant",
#                     "content": structured_response.punny_response
#                 },
#                 "finish_reason": "stop"
#             }
#         ]
#     }

# # ------------------------
# # 调用 Agent
# # ------------------------
# resp1 = agent.invoke(
#     {"messages": [{"role": "user", "content": "what is the weather outside?"}]},
#     config=config,
#     context=Context(user_id="1")
# )

# resp2 = agent.invoke(
#     {"messages": [{"role": "user", "content": "thank you!"}]},
#     config=config,
#     context=Context(user_id="1")
# )

# # ------------------------
# # 打印 OpenAI 风格输出
# # ------------------------
# print(wrap_openai_format(resp1['structured_response']))
# print(wrap_openai_format(resp2['structured_response']))



from dataclasses import dataclass
from langchain_openai import ChatOpenAI
from langgraph.checkpoint.memory import InMemorySaver
from langchain.agents.structured_output import ToolStrategy
from langchain.agents import create_agent
from langchain.tools import tool, ToolRuntime
import time

# ------------------------
# 系统提示（中文）
# ------------------------
SYSTEM_PROMPT = """
你是一位天气预报专家，说话风格幽默带双关。
你可以使用两个工具：
- get_weather_for_location(city)：查询指定城市的天气
- get_user_location()：获取用户所在城市
请回答时严格遵守 ResponseFormat 结构。
"""

# ------------------------
# 上下文
# ------------------------
@dataclass
class Context:
    user_id: str  # 用户 ID

# ------------------------
# 输出格式
# ------------------------
@dataclass
class ResponseFormat:
    punny_response: str           # 带双关语的回答
    weather_conditions: str | None = None  # 天气信息（可选）

# ------------------------
# 内存
# ------------------------
checkpointer = InMemorySaver()

# ------------------------
# 模型
# ------------------------
model = ChatOpenAI(
    base_url="http://localhost:40050/v1",
    model="Qwen/Qwen2-VL-7B-Instruct",
    temperature=0.1,
    api_key="EMPTY"
)

# ------------------------
# 工具函数
# ------------------------
@tool
def get_weather_for_location(city: str) -> str:
    """查询指定城市的天气"""
    return f"{city} 的天气总是晴朗！"

@tool
def get_user_location(runtime: ToolRuntime[Context]) -> str:
    """根据用户 ID 获取所在城市"""
    return "佛罗里达" if runtime.context.user_id == "1" else "旧金山"

# ------------------------
# 创建 Agent
# ------------------------
agent = create_agent(
    model=model,
    system_prompt=SYSTEM_PROMPT,
    tools=[get_user_location, get_weather_for_location],
    context_schema=Context,
    response_format=ToolStrategy(ResponseFormat),
    checkpointer=checkpointer
)

# ------------------------
# 配置线程
# ------------------------
config = {"configurable": {"thread_id": "1"}}

# ------------------------
# 辅助函数：包装成 OpenAI 风格 JSON
# ------------------------
def wrap_openai_format(structured_response: ResponseFormat):
    return {
        "id": f"chatcmpl-{int(time.time())}",
        "object": "chat.completion",
        "created": int(time.time()),
        "model": "Qwen/Qwen2-VL-7B-Instruct",
        "choices": [
            {
                "index": 0,
                "message": {
                    "role": "assistant",
                    "content": structured_response.punny_response
                },
                "finish_reason": "stop"
            }
        ]
    }

# ------------------------
# 多轮调用示例
# ------------------------
resp1 = agent.invoke(
    {"messages": [{"role": "user", "content": "现在外面天气怎么样？"}]},
    config=config,
    context=Context(user_id="1")
)
print(wrap_openai_format(resp1['structured_response']))

resp2 = agent.invoke(
    {"messages": [{"role": "user", "content": "谢谢你！"}]},
    config=config,
    context=Context(user_id="1")
)
print(wrap_openai_format(resp2['structured_response']))

---

openai.BadRequestError: Error code: 400 - {'error': {'message': "1 validation error for list[function-wrap[__log_extra_fields__()]]\n  Invalid JSON: EOF while parsing a list at line 2673 column 0 [type=json_invalid, input_value='[  \\n     \\n     \\n    \\...  \\n    \\n     \\n    \\n', input_type=str]\n    For further information visit https://errors.pydantic.dev/2.12/v/json_invalid", 'type': 'BadRequestError', 'param': None, 'code': 400}}