-
Notifications
You must be signed in to change notification settings - Fork 61.2k
New issue
Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? # to your account
docker部署的版本,用ngnix走https后GPT不会逐字回答 #209
Comments
参考 试试 |
感谢我会尝试一下! |
已解决,就是proxy_buffering off;的问题,取消缓存现在https反代正常使用! |
大哥,你是用docker部署的吗?为什么我部署会失败啊,就是代理问题,本地使用那个代理没问题,但是用docker部署的话就会报错 |
This was referenced Apr 23, 2023
Closed
ryiann
pushed a commit
to ryiann/ChatGPT
that referenced
this issue
Dec 28, 2023
- [+] refactor(model-config.tsx): remove unused code for input template and history count
# for free
to join this conversation on GitHub.
Already have an account?
# to comment
如题,现在我是用ngnix将3000端口映射到100端口绑定域名,并开启ssl,但是GPT回答不是正常一个字一个字出,而是等一会直接出完,并且说两句后会报错

但是如果直接通过IP或者域名访问3000端口就是正常的,问问大佬们,我该怎么解决,感谢!
The text was updated successfully, but these errors were encountered: