【GPT入门】第20课 langchain的function calling 初步体验

1 langchain的function calling 非常简洁

1. 在方法名说明方法用途和参数作用
2. 增加@tool 标签
3. langchain方法自动把@tool转为方法定义，后续方法调用都很简洁 下面代码用支持单函数与多函数调用，自己体验一下 query = “3 加 4的和 的 5 倍是多少？” 这个会调用两次函数 query = " 4的 5 倍是多少？"

2 代码

from langchain_core.messages import HumanMessage from langchain_core.tools import tool from langchain_openai import ChatOpenAI @tool def add(a: int, b: int) -> int: “““Add two integers together. Args: a: The first integer. b: The second integer. "”” return a + b @tool def multiply(a: int, b: int) -> int: “““Multiply two integers together. Args: a: The first integer. b: The second integer. "”” return a * b import json

model_name = ‘gpt-4o-mini’

model_name = ‘gpt-4’

llm = ChatOpenAI(model_name=model_name)

llm = ChatOpenAI() llm_with_tools = llm.bind_tools([add, multiply]) query = “3 加 4的和 的 5 倍是多少？”

query = " 4的 5 倍是多少？”

messages = [HumanMessage(content=query)] output = llm_with_tools.invoke(messages) print(output) print(”————tool_calls:") print(json.dumps(output.tool_calls, indent=4)) messages.append(output) avaliable_tools = {“add”:add, “multiply”: multiply} for tool_call in output.tool_calls: selected_tools = avaliable_tools[tool_call[“name”].lower()] # tool_call就是一个字典 tool_msg = selected_tools.invoke(tool_call) messages.append(tool_msg) print(“tool_msg:, type:”, type(tool_msg)) # print(tool_msg)

content=‘20’ name=‘multiply’ tool_call_id=‘call_e5EY7klNZlBD8W68y7X0BcYD’

new_output = llm_with_tools.invoke(messages) for message in messages: print(“message.dump:”) print(json.dumps(message.model_dump(), indent=4, ensure_ascii=False)) print(new_output.content)

3 执行结果：

C:\ProgramData\anaconda3\envs\gptLearning\python.exe E:\workspace\gptLearning\gptLearning\ls10\functionCalling\ls01_tools.py content=’’ additional_kwargs={’tool_calls’: [{‘id’: ‘call_0HGpTdeo2zm9G6zZPbewVi8g’, ‘function’: {‘arguments’: ‘{“a”: 3, “b”: 4}’, ’name’: ‘add’}, ’type’: ‘function’}, {‘id’: ‘call_m6Uc0OfQ9ea0uTyFNbwEWIBq’, ‘function’: {‘arguments’: ‘{“a”: 7, “b”: 5}’, ’name’: ‘multiply’}, ’type’: ‘function’}], ‘refusal’: None} response_metadata={’token_usage’: {‘completion_tokens’: 49, ‘prompt_tokens’: 121, ’total_tokens’: 170, ‘completion_tokens_details’: None, ‘prompt_tokens_details’: None}, ‘model_name’: ‘gpt-3.5-turbo-0125’, ‘system_fingerprint’: ‘fp_0165350fbb’, ‘finish_reason’: ’tool_calls’, ’logprobs’: None} id=‘run-e83caecd-ca6f-4851-b35e-b5ea8f2804a2-0’ tool_calls=[{’name’: ‘add’, ‘args’: {‘a’: 3, ‘b’: 4}, ‘id’: ‘call_0HGpTdeo2zm9G6zZPbewVi8g’, ’type’: ’tool_call’}, {’name’: ‘multiply’, ‘args’: {‘a’: 7, ‘b’: 5}, ‘id’: ‘call_m6Uc0OfQ9ea0uTyFNbwEWIBq’, ’type’: ’tool_call’}] usage_metadata={‘input_tokens’: 121, ‘output_tokens’: 49, ’total_tokens’: 170, ‘input_token_details’: {}, ‘output_token_details’: {}} ————tool_calls: [ { “name”: “add”, “args”: { “a”: 3, “b”: 4 }, “id”: “call_0HGpTdeo2zm9G6zZPbewVi8g”, “type”: “tool_call” }, { “name”: “multiply”, “args”: { “a”: 7, “b”: 5 }, “id”: “call_m6Uc0OfQ9ea0uTyFNbwEWIBq”, “type”: “tool_call” } ] tool_msg:, type: content=‘7’ name=‘add’ tool_call_id=‘call_0HGpTdeo2zm9G6zZPbewVi8g’ tool_msg:, type: content=‘35’ name=‘multiply’ tool_call_id=‘call_m6Uc0OfQ9ea0uTyFNbwEWIBq’ message.dump: { “content”: “3 加 4的和 的 5 倍是多少？”, “additional_kwargs”: {}, “response_metadata”: {}, “type”: “human”, “name”: null, “id”: null, “example”: false } message.dump: { “content”: “”, “additional_kwargs”: { “tool_calls”: [ { “id”: “call_0HGpTdeo2zm9G6zZPbewVi8g”, “function”: { “arguments”: “{"a": 3, "b": 4}”, “name”: “add” }, “type”: “function” }, { “id”: “call_m6Uc0OfQ9ea0uTyFNbwEWIBq”, “function”: { “arguments”: “{"a": 7, "b": 5}”, “name”: “multiply” }, “type”: “function” } ], “refusal”: null }, “response_metadata”: { “token_usage”: { “completion_tokens”: 49, “prompt_tokens”: 121, “total_tokens”: 170, “completion_tokens_details”: null, “prompt_tokens_details”: null }, “model_name”: “gpt-3.5-turbo-0125”, “system_fingerprint”: “fp_0165350fbb”, “finish_reason”: “tool_calls”, “logprobs”: null }, “type”: “ai”, “name”: null, “id”: “run-e83caecd-ca6f-4851-b35e-b5ea8f2804a2-0”, “example”: false, “tool_calls”: [ { “name”: “add”, “args”: { “a”: 3, “b”: 4 }, “id”: “call_0HGpTdeo2zm9G6zZPbewVi8g”, “type”: “tool_call” }, { “name”: “multiply”, “args”: { “a”: 7, “b”: 5 }, “id”: “call_m6Uc0OfQ9ea0uTyFNbwEWIBq”, “type”: “tool_call” } ], “invalid_tool_calls”: [], “usage_metadata”: { “input_tokens”: 121, “output_tokens”: 49, “total_tokens”: 170, “input_token_details”: {}, “output_token_details”: {} } } message.dump: { “content”: “7”, “additional_kwargs”: {}, “response_metadata”: {}, “type”: “tool”, “name”: “add”, “id”: null, “tool_call_id”: “call_0HGpTdeo2zm9G6zZPbewVi8g”, “artifact”: null, “status”: “success” } message.dump: { “content”: “35”, “additional_kwargs”: {}, “response_metadata”: {}, “type”: “tool”, “name”: “multiply”, “id”: null, “tool_call_id”: “call_m6Uc0OfQ9ea0uTyFNbwEWIBq”, “artifact”: null, “status”: “success” } 3 加 4 的和是 7，7 的 5 倍是 35。 Process finished with exit code 0 ``