Merge pull request xinnan-tech#962 from xinnan-tech/update-fix

update:让alibl、等非原生llm使用functional时不卡顿

Author: xinnan-tech

docs/firmware-build.md

@@ -1,16 +1,40 @@
 # esp32固件编译
 
-## 第1步 配置环境
+## 第1步 准备你的ota地址
+如果你按照教程使用的是全模块部署，就应该会有ota地址。
+
+此刻，请你用浏览器打开你的ota地址，例如我的ota地址
+```
+http://192.168.1.25:8002/xiaozhi/ota/
+```
+
+如果显示“OTA接口运行正常，websocket集群数量：X”。那就往下。
+
+如果显示“OTA接口运行不正常”，大概是你还没在`智控台`配置`Websocket`地址。那就：
+
+- 1、使用超级管理员登录智控台
+
+- 2、顶部菜单点击`参数管理`
+
+- 3、在列表中找到`server.websocket`项目，输入你的`Websocket`地址。例如我的就是
+
+```
+ws://192.168.1.25:8000/xiaozhi/v1/
+```
+
+配置完后，再使用浏览器刷新你的ota接口地址，看看是不是正常了。如果还不正常就，就再次确认一下Websocket是否正常启动，是否配置了Websocket地址。
+
+## 第2步 配置环境
 先按照这个教程配置项目环境[《Windows搭建 ESP IDF 5.3.2开发环境以及编译小智》](https://icnynnzcwou8.feishu.cn/wiki/JEYDwTTALi5s2zkGlFGcDiRknXf)
 
-## 第2步 打开配置文件
+## 第3步 打开配置文件
 配置好编译环境后，下载虾哥iaozhi-esp32项目源码，
 
 从这里下载虾哥[xiaozhi-esp32项目源码](https://github.com/78/xiaozhi-esp32)。
 
 下载后，打开`xiaozhi-esp32/main/Kconfig.projbuild`文件。
 
-## 第3步 修改OTA地址
+## 第4步 修改OTA地址
 
 找到`OTA_VERSION_URL`的`default`的内容，把`https://api.tenclass.net/xiaozhi/ota/`
    改成你自己的地址，例如，我的接口地址是`http://192.168.1.25:8002/xiaozhi/ota/`，就把内容改成这个。

main/xiaozhi-server/core/providers/llm/AliBL/AliBL.py

@@ -6,6 +6,7 @@
 TAG = __name__
 logger = setup_logging()
 
+
 class LLMProvider(LLMProviderBase):
     def __init__(self, config):
         self.api_key = config["api_key"]
@@ -19,21 +20,25 @@ def response(self, session_id, dialogue):
             # 处理dialogue
             if self.is_No_prompt:
                 dialogue.pop(0)
-                logger.bind(tag=TAG).debug(f"【阿里百练API服务】处理后的dialogue: {dialogue}")
+                logger.bind(tag=TAG).debug(
+                    f"【阿里百练API服务】处理后的dialogue: {dialogue}"
+                )
 
             # 构造调用参数
             call_params = {
                 "api_key": self.api_key,
                 "app_id": self.app_id,
                 "session_id": session_id,
-                "messages": dialogue
+                "messages": dialogue,
             }
             if self.memory_id != False:
                 # 百练memory需要prompt参数
                 prompt = dialogue[-1].get("content")
                 call_params["memory_id"] = self.memory_id
                 call_params["prompt"] = prompt
-                logger.bind(tag=TAG).debug(f"【阿里百练API服务】处理后的prompt: {prompt}")
+                logger.bind(tag=TAG).debug(
+                    f"【阿里百练API服务】处理后的prompt: {prompt}"
+                )
 
             responses = Application.call(**call_params)
             if responses.status_code != HTTPStatus.OK:
@@ -44,9 +49,15 @@ def response(self, session_id, dialogue):
                 )
                 yield "【阿里百练API服务响应异常】"
             else:
-                logger.bind(tag=TAG).debug(f"【阿里百练API服务】构造参数: {call_params}")
+                logger.bind(tag=TAG).debug(
+                    f"【阿里百练API服务】构造参数: {call_params}"
+                )
                 yield responses.output.text
 
         except Exception as e:
             logger.bind(tag=TAG).error(f"【阿里百练API服务】响应异常: {e}")
             yield "【LLM服务响应异常】"
+
+    def response_with_functions(self, session_id, dialogue, functions=None):
+        logger.bind(tag=TAG).info(f"阿里百练暂未实现完整的工具调用（function call）")
+        return self.response(session_id, dialogue)

main/xiaozhi-server/core/providers/llm/fastgpt/fastgpt.py

@@ -21,37 +21,36 @@ def response(self, session_id, dialogue):
 
             # 发起流式请求
             with requests.post(
-                    f"{self.base_url}/chat/completions",
-                    headers={"Authorization": f"Bearer {self.api_key}"},
-                    json={
-                        "stream": True,
-                        "chatId": session_id,
-                        "detail": self.detail,
-                        "variables": self.variables,
-                        "messages": [
-                            {
-                                "role": "user",
-                                "content": last_msg["content"]
-                            }
-                        ]
-                    },
-                    stream=True
+                f"{self.base_url}/chat/completions",
+                headers={"Authorization": f"Bearer {self.api_key}"},
+                json={
+                    "stream": True,
+                    "chatId": session_id,
+                    "detail": self.detail,
+                    "variables": self.variables,
+                    "messages": [{"role": "user", "content": last_msg["content"]}],
+                },
+                stream=True,
             ) as r:
                 for line in r.iter_lines():
                     if line:
                         try:
-                            if line.startswith(b'data: '):
-                                if line[6:].decode('utf-8') == '[DONE]':
+                            if line.startswith(b"data: "):
+                                if line[6:].decode("utf-8") == "[DONE]":
                                     break
 
                                 data = json.loads(line[6:])
-                                if 'choices' in data and len(data['choices']) > 0:
-                                    delta = data['choices'][0].get('delta', {})
-                                    if delta and 'content' in delta and delta['content'] is not None:
-                                        content = delta['content']
-                                        if '<think>' in content:
+                                if "choices" in data and len(data["choices"]) > 0:
+                                    delta = data["choices"][0].get("delta", {})
+                                    if (
+                                        delta
+                                        and "content" in delta
+                                        and delta["content"] is not None
+                                    ):
+                                        content = delta["content"]
+                                        if "<think>" in content:
                                             continue
-                                        if '</think>' in content:
+                                        if "</think>" in content:
                                             continue
                                         yield content
 
@@ -62,4 +61,8 @@ def response(self, session_id, dialogue):
 
         except Exception as e:
             logger.bind(tag=TAG).error(f"Error in response generation: {e}")
-            yield "【服务响应异常】"
+            yield "【服务响应异常】"
+
+    def response_with_functions(self, session_id, dialogue, functions=None):
+        logger.bind(tag=TAG).info(f"fastgpt暂未实现完整的工具调用（function call）")
+        return self.response(session_id, dialogue)

main/xiaozhi-server/core/providers/llm/gemini/gemini.py

@@ -134,3 +134,7 @@ def response(self, session_id, dialogue):
             yield f"JSON解码错误：{e}"
         except Exception as e:
             yield f"发生错误：{e}"
+
+    def response_with_functions(self, session_id, dialogue, functions=None):
+        logger.bind(tag=TAG).info(f"gemini暂未实现完整的工具调用（function call）")
+        return self.response(session_id, dialogue)