Merge pull request #312 from mawwalker/master

Feat: OpenAIrobot类增加微软azure支持

robot/AI.py

@@ -226,7 +226,6 @@ def chat(self, texts, parsed):
             logger.critical("AnyQ robot failed to response for %r", msg, exc_info=True)
             return "抱歉, AnyQ回答失败"
 
-
 class OPENAIRobot(AbstractRobot):
 
     SLUG = "openai"
@@ -235,6 +234,8 @@ def __init__(
         self,
         openai_api_key,
         model,
+        provider,
+        api_version,
         temperature,
         max_tokens,
         top_p,
@@ -267,6 +268,8 @@ def __init__(
             logger.critical("OpenAI 初始化失败，请升级 Python 版本至 > 3.6")
         self.model = model
         self.prefix = prefix
+        self.provider = provider
+        self.api_version = api_version
         self.temperature = temperature
         self.max_tokens = max_tokens
         self.top_p = top_p
@@ -295,12 +298,20 @@ def stream_chat(self, texts):
 
         header = {
             "Content-Type": "application/json",
-            "Authorization": "Bearer " + self.openai.api_key,
+            # "Authorization": "Bearer " + self.openai.api_key,
         }
+        if self.provider == 'openai':
+            header['Authorization'] = "Bearer " + self.openai.api_key,
+        elif self.provider == 'azure':
+            header['api-key'] = self.openai.api_key
+        else:
+            raise ValueError("Please check your config file, OpenAiRobot's provider should be openai or azure.")
 
         data = {"model": self.model, "messages": self.context, "stream": True}
         logger.info(f"使用模型：{self.model}，开始流式请求")
         url = self.api_base + "/completions"
+        if self.provider == 'azure':
+            url = f"{self.api_base}/openai/deployments/{self.model}/chat/completions?api-version={self.api_version}"
         # 请求接收流式数据
         try:
             response = requests.request(
@@ -368,17 +379,29 @@ def chat(self, texts, parsed):
         try:
             respond = ""
             self.context.append({"role": "user", "content": msg})
-            response = self.openai.Completion.create(
-                model=self.model,
-                messages=self.context,
-                temperature=self.temperature,
-                max_tokens=self.max_tokens,
-                top_p=self.top_p,
-                frequency_penalty=self.frequency_penalty,
-                presence_penalty=self.presence_penalty,
-                stop=self.stop_ai,
-                api_base=self.api_base
-            )
+            if self.provider == "openai":
+                response = self.openai.Completion.create(
+                    model=self.model,
+                    messages=self.context,
+                    temperature=self.temperature,
+                    max_tokens=self.max_tokens,
+                    top_p=self.top_p,
+                    frequency_penalty=self.frequency_penalty,
+                    presence_penalty=self.presence_penalty,
+                    stop=self.stop_ai,
+                    api_base=self.api_base
+                )
+            else:
+                from openai import AzureOpenAI
+                client = AzureOpenAI(
+                            azure_endpoint = self.api_base, 
+                            api_key=self.openai_api_key,  
+                            api_version=self.api_version
+                            )
+                response = client.chat.completions.create(
+                    model=self.model,
+                    messages=self.context
+                )
             message = response.choices[0].message
             respond = message.content
             self.context.append(message)

static/default.yml

@@ -284,6 +284,8 @@ tuling:
 # 注册一个账号，获得 openai_api_key 后填到下面的配置中即可
 openai:
     openai_api_key: 'sk-xxxxxxxxxxxxxxxxxxxxxxxxxx'
+    provider: 'azure' # openai的接口填写openai, azure的填写azure
+    api_version: '2023-05-15' # 如果是openai的，留空就行，azure的需填写对应的api_version，参考官方文档
     # 参数指定将生成文本的模型类型。目前支持 gpt-3.5-turbo 和 gpt-3.5-turbo-0301 两种选择
     model: 'gpt-3.5-turbo'
     # 在前面加的一段前缀