目前出了新模型了，项目是否已经支持啦？需要更改config么？

[dwd923]

⚠️ 搜索是否存在类似issue

• [X] 我已经搜索过issues和disscussions，没有发现相似issue

总结

No response

举例

No response

动机

No response

[lanvent]

修改配置中的model就能使用，如果要使用16k model，conversation_max_tokens 需要配合修改大一点

[dwd923]

修改配置中的model就能使用，如果要使用16k model，conversation_max_tokens 需要配合修改大一点

修改成gpt-3.5-turbo-16k-0613是不是就可以？

[lanvent]

• gpt3.5 "gpt-3.5-turbo-0301", "gpt-3.5-turbo-0613", "gpt-35-turbo", "gpt-3.5-turbo-16k", "gpt-3.5-turbo-16k-0613"
• gpt4 "gpt-4-0314", "gpt-4-0613", "gpt-4-32k", "gpt-4-32k-0613"

[dwd923]

get√，谢大佬

[GesiIvy128]

我将model改为"gpt-3.5-turbo-0613"提问后却报错： [INFO][2023-06-17 10:55:28][chat_gpt_bot.py:49] - [CHATGPT] query=你在吗 [WARNING][2023-06-17 10:55:30][chat_gpt_session.py:77] - num_tokens_from_messages() is not implemented for model gpt-3.5-turbo-0613. Returning num tokens assuming gpt-3.5-turbo-0301. [WARNING][2023-06-17 10:55:31][chat_gpt_session.py:77] - num_tokens_from_messages() is not implemented for model gpt-3.5-turbo-0613. Returning num tokens assuming gpt-3.5-turbo-0301. [ERROR][2023-06-17 10:55:31][chat_channel.py:270] - Worker return exception: can only concatenate list (not "str") to list Traceback (most recent call last): File "/usr/local/lib/python3.10/concurrent/futures/thread.py", line 58, in run result = self.fn(*self.args, **self.kwargs) File "/app/channel/chat_channel.py", line 150, in _handle reply = self._decorate_reply(context, reply) File "/app/channel/chat_channel.py", line 228, in _decorate_reply reply_text = conf().get("single_chat_reply_prefix", "") + reply_text TypeError: can only concatenate list (not "str") to list

我是用railway部署的

[chenjin0827]

一样的问题,解决了吗

[shxch]

相同的问题，我觉得应该把单聊prefix去掉，应该没人会在单聊的时候还加个前缀吧

[INFO][2023-07-12 18:17:23][chat_gpt_bot.py:49] - [CHATGPT] query=hi [ERROR][2023-07-12 18:17:26][chat_channel.py:270] - Worker return exception: can only concatenate list (not "str") to list Traceback (most recent call last): File "/usr/local/lib/python3.10/concurrent/futures/thread.py", line 58, in run result = self.fn(*self.args, **self.kwargs) File "/app/channel/chat_channel.py", line 150, in _handle reply = self._decorate_reply(context, reply) File "/app/channel/chat_channel.py", line 228, in _decorate_reply reply_text = conf().get("single_chat_reply_prefix", "") + reply_text + conf().get("single_chat_reply_suffix", "") TypeError: can only concatenate list (not "str") to list

[GesiIvy128]

我觉得从逻辑角度 不该去掉单聊的prefix, 因为你的小号可能加了其他人，你总不会希望随便一个什么人问你小号问题，你的小号都化身bot吧

相同的问题，我觉得应该把单聊prefix去掉，应该没人会在单聊的时候还加个前缀吧

[INFO][2023-07-12 18:17:23][chat_gpt_bot.py:49] - [CHATGPT] query=hi [ERROR][2023-07-12 18:17:26][chat_channel.py:270] - Worker return exception: can only concatenate list (not "str") to list Traceback (most recent call last): File "/usr/local/lib/python3.10/concurrent/futures/thread.py", line 58, in run result = self.fn(*self.args, **self.kwargs) File "/app/channel/chat_channel.py", line 150, in _handle reply = self._decorate_reply(context, reply) File "/app/channel/chat_channel.py", line 228, in _decorate_reply reply_text = conf().get("single_chat_reply_prefix", "") + reply_text + conf().get("single_chat_reply_suffix", "") TypeError: can only concatenate list (not "str") to list

[shxch]

有道理诶，是我考虑不周了

[hellosmarter]

@lanvent 修改配置中的model修改成了gpt-4-0613，运行是提示下面这个告警，这个应该不影响使用gpt4模型吧，只是说这个token计算不准是吧？ 另外最新的代码有没有把这个警告fix掉了，因为我用的还是比较之前旧一点的代码，谢谢 [WARNING][2023-07-31 00:12:48][chat_gpt_session.py:76] - num_tokens_from_messages() is not implemented for model gpt-4-0613. Returning num tokens assuming gpt-3.5-turbo-0301.