1.找到knowledge_base_chat.py文件中的get_ChatOpenAI函数
2.按crtl进入get_ChatOpenAI函数位置
3.注释原先的get_ChatOpenAI函数,修改成以下内容:
def get_ChatOpenAI(
model_name: str,
temperature: float,
streaming: bool = True,
callbacks: List[Callable] = [],
verbose: bool = True,
**kwargs: Any,
) -> AzureChatOpenAI:
config = get_model_worker_config(model_name)
# if model_name == "openai-api":
# model_name = config.get("model_name")
model = AzureChatOpenAI(
streaming=streaming,
verbose=verbose,
callbacks=callbacks,
openai_api_key=config.get("api_key"),
openai_api_base=config.get("api_base_url"),
deployment_name=config.get("model_name"),
temperature=temperature,
# openai_proxy=config.get("openai_proxy"),
openai_api_version="2024-02-01",
# openai_api_type="azure",
** kwargs
)
return model
4.找到model_config.py文件
5.增加以下内容
(1)创建.env文件,里面放入自己在微软申请的Azure Open AI
(2)加载.env文件,在model_config.py文件加上以下代码
from dotenv import load_dotenv
load_dotenv()
(3)修改openai-api,在model_config.py文件加上以下代码
ONLINE_LLM_MODEL = {
"openai-api": {
"model_name": "otx_dmed_gpt4",
"api_base_url": os.getenv("AZURE_OAI_ENDPOINT"),
"api_key": os.getenv("AZURE_OAI_KEY"),
"openai_proxy": "",
},
6.直接python startup.py -a运行即可调用Azure Open AI