[Bug] 在使用Upstash同步时，模型会自动恢复gpt-3.5-turbo

[aaro-n]

Describe the bug 在ChatGPT-Next-Web配置界面，配置Upstash同步，同步成功后，在同配置界面选择模型：gpt-4，再点击同步按钮此时 模型会变成gpt-3.5-turbo，无论怎么改都无法生效，如果在单独对话框选择gpt-4，同步后这个对话还是gpt4，单独选择是正常的。

To Reproduce Steps to reproduce the behavior:

1. 配置Upstash同步，并测试同步，此时模型是gpt-3.5-turbo
2. 修改模型为gpt-4，再次点击同步，此时模型会由gpt-4自动变为gpt-3.5-turbo
3. 如果在对话框选择gpt-4，同步后，这个对话框还是gpt-4，这个正常。

Expected behavior 选择的默认模型能修改后成功同步

Screenshots If applicable, add screenshots to help explain your problem.

Deployment

• [ ] Docker
• [x] Vercel
• [ ] Server

Desktop (please complete the following information):

• OS: windows 10 、macos
• Browser chrome
• Version chrome 版本 118.0.5993.70（正式版本） (x86_64)

Additional Logs Add any logs about the problem here.

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

Title: [Bug] When using Upstash to synchronize, the model will automatically restore gpt-3.5-turbo

[aaro-n]

在配置界面设置发送键时，同步后还是会还原为Enter，而我想要的是Ctrl+Enter.

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

When setting the Send key in the configuration interface, it will still be restored to Enter after synchronization, and what I want is Ctrl+Enter.

[Cassianvale]

Currently, when using the synchronization feature, historical chats and settings will not be deleted. They will only be added to the existing content. In the future, there may be additional constraint rules implemented.

[hqw567]

I had the same problem. @Yidadaa Upstash only synchronizes configuration parameters for the first time, and only synchronizes chat records afterwards, but does not synchronize configuration parameters, resulting in configuration parameters that are old before each synchronization.

[shaolinah]

这里可以指定，同步时也不会覆盖

[Issues-translate-bot]

Bot detected the issue body's language is not English, translate it automatically.

---

You can specify it here and it will not be overwritten during synchronization.