langchain接入各种模型

langchain适配的各种大模型可能会不能适配各家最新版本，这时候需要自己去封装

智谱ai的Chat models

智谱官方调用api

官方控制台

我这里选用的是免费的GLM-4-FLASH

接口文档

设置环境变量或者.env中

环境变量的名字是自定义的,我这里为了后续跟langchain对齐，使用的名字为langchain中定义的

ZHIPUAI_API_KEY=

安装最新版python sdk

pip install --upgrade zhipuai

from zhipuai import ZhipuAI
import os
from dotenv import load_dotenv, find_dotenv
_ = load_dotenv(find_dotenv(), verbose=True)

zhipuai_api_key = os.getenv("ZHIPUAI_API_KEY")
client = ZhipuAI(api_key = zhipuai_api_key)

prompt = "为什么天空是蓝色的？"

response = client.chat.completions.create(
  model = "glm-4-flash",
  messages = [
    {"role": "user", "content": "你好"},
    {"role": "assistant", "content": "你好，我是气象学家"},
    {"role": "user", "content": prompt}
  ]
)

print(response.choices[0].message)

langchain适配的

langchain关于智谱ai的文档

安装依赖

pip install --upgrade zhipuai
# 因为zhipuai依赖 PyJWT版本为2.8-2.9
pip uninstall PyJWT
pip install PyJWT==2.8
pip install --upgrade httpx httpx-sse
pip install --upgrade langchain-community langchain-core

from langchain_community.chat_models import ChatZhipuAI
from langchain_core.messages import AIMessage, HumanMessage, SystemMessage
from dotenv import load_dotenv,find_dotenv

_ = load_dotenv(find_dotenv(), verbose=True)

chat = ChatZhipuAI(
  model="glm-4-flash",
  temperature=0.5,
)
messages = [
    AIMessage(content="Hi."),
    SystemMessage(content="Your role is a poet."),
    HumanMessage(content="Write a short poem about AI in four lines."),
]

response = chat.invoke(messages)
print(response.content)  # Displays the AI-generated poem