Merge pull request #2300 from 6vision/master

feat: support o1-preview and o1-mini model
zhayujie · Sep 13, 2024 · a11d695 · a11d695 · Admin-Author · Sep 13, 2024
2 parents bb5ecfc + c4f9acd
commit a11d695
Show file tree

Hide file tree

Showing 2 changed files with 16 additions and 5 deletions.
diff --git a/bot/chatgpt/chat_gpt_bot.py b/bot/chatgpt/chat_gpt_bot.py
@@ -5,7 +5,7 @@
 import openai
 import openai.error
 import requests
-
+from common import const
 from bot.bot import Bot
 from bot.chatgpt.chat_gpt_session import ChatGPTSession
 from bot.openai.open_ai_image import OpenAIImage
@@ -15,7 +15,7 @@
 from common.log import logger
 from common.token_bucket import TokenBucket
 from config import conf, load_config
-
+from bot.baidu.baidu_wenxin_session import BaiduWenxinSession
 
 # OpenAI对话模型API (可用)
 class ChatGPTBot(Bot, OpenAIImage):
@@ -30,10 +30,12 @@ def __init__(self):
             openai.proxy = proxy
         if conf().get("rate_limit_chatgpt"):
             self.tb4chatgpt = TokenBucket(conf().get("rate_limit_chatgpt", 20))
-
+        conf_model = conf().get("model") or "gpt-3.5-turbo"
         self.sessions = SessionManager(ChatGPTSession, model=conf().get("model") or "gpt-3.5-turbo")
+        # o1相关模型不支持system prompt，暂时用文心模型的session
+
         self.args = {
-            "model": conf().get("model") or "gpt-3.5-turbo",  # 对话模型的名称
+            "model": conf_model,  # 对话模型的名称
             "temperature": conf().get("temperature", 0.9),  # 值在[0,1]之间，越大表示回复越具有不确定性
             # "max_tokens":4096,  # 回复最大的字符数
             "top_p": conf().get("top_p", 1),
@@ -42,6 +44,12 @@ def __init__(self):
             "request_timeout": conf().get("request_timeout", None),  # 请求超时时间，openai接口默认设置为600，对于难问题一般需要较长时间
             "timeout": conf().get("request_timeout", None),  # 重试超时时间，在这个时间内，将会自动重试
         }
+        # o1相关模型固定了部分参数，暂时去掉
+        if conf_model in [const.O1, const.O1_MINI]:
+            self.sessions = SessionManager(BaiduWenxinSession, model=conf().get("model") or const.O1_MINI)
+            remove_keys = ["temperature", "top_p", "frequency_penalty", "presence_penalty"]
+            for key in remove_keys:
+                self.args.pop(key, None)  # 如果键不存在，使用 None 来避免抛出错误
 
     def reply(self, query, context=None):
         # acquire reply content

diff --git a/common/const.py b/common/const.py
@@ -38,6 +38,9 @@
 GPT4_06_13 = "gpt-4-0613"
 GPT4_32k_06_13 = "gpt-4-32k-0613"
 
+O1 = "o1-preview"
+O1_MINI = "o1-mini"
+
 WHISPER_1 = "whisper-1"
 TTS_1 = "tts-1"
 TTS_1_HD = "tts-1-hd"
@@ -59,7 +62,7 @@
 
 MODEL_LIST = [
               GPT35, GPT35_0125, GPT35_1106, "gpt-3.5-turbo-16k",
-              GPT_4o, GPT_4O_0806, GPT_4o_MINI, GPT4_TURBO, GPT4_TURBO_PREVIEW, GPT4_TURBO_01_25, GPT4_TURBO_11_06, GPT4, GPT4_32k, GPT4_06_13, GPT4_32k_06_13,
+              O1, O1_MINI, GPT_4o, GPT_4O_0806, GPT_4o_MINI, GPT4_TURBO, GPT4_TURBO_PREVIEW, GPT4_TURBO_01_25, GPT4_TURBO_11_06, GPT4, GPT4_32k, GPT4_06_13, GPT4_32k_06_13,
               WEN_XIN, WEN_XIN_4,
               XUNFEI, ZHIPU_AI, MOONSHOT, MiniMax,
               GEMINI, GEMINI_PRO, GEMINI_15_flash, GEMINI_15_PRO,