ChatGPTNextWeb
[Bug] 新版本的DEFAULT_MODEL怎么设置呢？
📦 部署方式
Docker
📌 软件版本
V2.14
💻 系统环境
Windows
📌 系统版本
Win 11
🌐 浏览器
Chrome
📌 浏览器版本
最新
🐛 问题描述
原来V2.12版本设置默认核心只需在Docker部署时加上：-e DEFAULT_MODEL=gemini-1.5-pro-latest，V2.13以后就不行了，另外设置以下格式也不行：-e DEFAULT_MODEL=gemini-1.5-pro-latest@Google ，默认核心一直是：gpt-3.5-turbo(OpenAi)
📷 复现步骤
No response
🚦 期望结果
No response
📝 补充信息
No response
Bot detected the issue body's language is not English, translate it automatically.
Title: [Bug] How to set the new version of DEFAULT_MODEL?
📦 Deployment method
Docker
📌 Software version
V2.14
💻 System environment
Windows
📌 System version
Win 11
🌐 Browser
Chrome
📌 Browser version
up to date
🐛 Problem description
It turns out that to set the default core in version V2.12, you only need to add: -e DEFAULT_MODEL=gemini-1.5-pro-latest when deploying Docker. This will not work after V2.13. In addition, setting the following format will not work: -e DEFAULT_MODEL=gemini- 1.5-pro-latest@Google, the default core has always been: gpt-3.5-turbo (OpenAi)
📷 Steps to reproduce
No response
🚦 Expected results
No response
📝 Supplementary information
No response
Bot detected the issue body's language is not English, translate it automatically.
There is this piece of code in the new version of ./app/config/server.ts:
It is speculated that this is the cause, but I don’t know the meaning of the author adding this code that ignores gpt-4.
新版本的./app/config/server.ts中有这样一段代码：
推测是这个导致的，但不知道作者加这段无视gpt-4的代码的意义
只有添加了 DISABLE_GPT4 环境变量才会使用到这部分的逻辑，用来排除所有 GPT4 的模型，似乎和这个ISSUES没有关系
Bot detected the issue body's language is not English, translate it automatically.
The new version of ./app/config/server.ts contains this piece of code:
It is speculated that this is the cause, but I don’t know the meaning of the author adding this code that ignores gpt-4.
This part of the logic will only be used if the DISABLE_GPT4 environment variable is added to exclude all GPT4 models. It seems to have nothing to do with this ISSUES.
Bot detected the issue body's language is not English, translate it automatically.
I also had this problem when I deployed using Vercel. It was set to 4o-mini but it never took effect. I redeployed it many times and it still didn't work.
-e DEFAULT_MODEL=gemini-1.5-pro-latest@google 以上参数设置可以使用，不过设置中的默认模型还是gpt-3.5-turbo
Bot detected the issue body's language is not English, translate it automatically.
-e DEFAULT_MODEL=gemini-1.5-pro-latest@google The above parameter settings can be used, but the default model in the settings is still gpt-3.5-turbo
Bot detected the issue body's language is not English, translate it automatically.
In the latest version, v2.14.2, setting DEFAULT_MODEL has no effect.
Bot detected the issue body's language is not English, translate it automatically.
My docker deployment is the same. Setting DEFAULT_MODEL is invalid, and it still displays gpt-3.5. Please fix it.
Bot detected the issue body's language is not English, translate it automatically.
This is indeed a bug, and the DEFAULT_MODEL I set does not take effect.
Bot detected the issue body's language is not English, translate it automatically.
fixed