ChatGPTNextWeb / ChatGPT-Next-Web

A cross-platform ChatGPT/Gemini UI (Web / PWA / Linux / Win / MacOS). 一键拥有你自己的跨平台 ChatGPT/Gemini 应用。
https://app.nextchat.dev/
MIT License
76.47k stars 59.13k forks source link

[Feature Request]: 设置了只有 gpt-4o,但是会显示 2 个版本,分别是 azure 和 openai最近一次更新才发现的。 #4940

Closed itcodes closed 4 months ago

itcodes commented 4 months ago

Problem Description

CleanShot 2024-07-07 at 18 49 55@2x

Solution Description

原来没出现过这个问题,是最近更新才遇到。

Alternatives Considered

No response

Additional Context

No response

Issues-translate-bot commented 4 months ago

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


Title: [Feature Request]: Only gpt-4o is set, but 2 versions will be displayed, which were discovered by the latest update of azure and openai.

Problem Description

CleanShot 2024-07-07 at 18 49 55@2x

Solution Description

This problem never occurred before, it was only encountered in the latest update.

Alternatives Considered

No response

Additional Context

No response

itcodes commented 4 months ago

CleanShot 2024-07-07 at 18 47 50@2x

Issues-translate-bot commented 4 months ago

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


CleanShot 2024-07-07 at 18 47 50@2x

madagastar commented 4 months ago

用vercel部署,自动更新遇到了同样的问题。

Issues-translate-bot commented 4 months ago

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


Deployed with vercel, automatic update encountered the same problem.

madagastar commented 4 months ago

屏幕截图 2024-07-08 131236

Issues-translate-bot commented 4 months ago

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


Screenshot 2024-07-08 131236

itcodes commented 4 months ago

用vercel部署,自动更新遇到了同样的问题。

对的。就是4930 那个 pr 引起的。下次不自动更新了。

Issues-translate-bot commented 4 months ago

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


Deploying with vercel, automatic update encountered the same problem.

Yes. It was caused by the PR of 4930.

lloydzhou commented 4 months ago

4945

fixed in #4953

manjieqi commented 4 months ago

这个重复的问题好像没有解决,vercel部署拉取的最新代码,-all,+gpt-3.5-turbo,+gpt-4,+gpt-4o,没有关于azure的环境变量,这样配置还是显示了azure的选项

Issues-translate-bot commented 4 months ago

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


This duplicate problem seems to have not been solved. The latest code pulled by vercel deployment, -all, +gpt-3.5-turbo, +gpt-4, +gpt-4o, does not have environment variables for azure. This configuration still shows azure. Options

lloydzhou commented 4 months ago

这个重复的问题好像没有解决,vercel部署拉取的最新代码,-all,+gpt-3.5-turbo,+gpt-4,+gpt-4o,没有关于azure的环境变量,这样配置还是显示了azure的选项

  1. 新的版本支持将azure完全独立出来作为一个供应商。
  2. 这个逻辑上的更改并不能完全的向前兼容,这里配置模型的时候,如果多个供应商提供相同的模型,需要在CUSTOM_MODELS里面显示的标记供应商名称(这里不做兼容的原因是,用户配置了并保存在localStorage中的自定义模型并不能区分究竟是azure还是openai提供) 以下是新的配置示例(readme中也有详细的说明):

    -all,+gpt-3.5-turbo@openai,+gpt-4@openai,+gpt-4o@openai -all,+gpt-3.5-turbo@azure={deploy_name} 这个配置会在模型列表展示一个“{deploy_name}(Azure)”的选项,发请求的时候使用的模型是gpt-3.5-turbo

Issues-translate-bot commented 4 months ago

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


This duplicate problem seems to have not been solved. The latest code pulled by vercel deployment, -all, +gpt-3.5-turbo, +gpt-4, +gpt-4o, does not have environment variables about azure. This configuration still displays azure. Options

  1. The new version supports Azure being completely independent as a supplier.
  2. This logical change is not completely forward compatible. When configuring the model here, if multiple suppliers provide the same model, the supplier name needs to be marked in CUSTOM_MODELS (the reason why compatibility is not done here is, The custom model configured by the user and saved in localStorage cannot distinguish whether it is provided by Azure or OpenAI) The following is a new configuration example (detailed instructions are also available in the readme):

    -all,+gpt-3.5-turbo@openai,+gpt-4@openai,+gpt-4o@openai -all,+gpt-3.5-turbo@azure={deploy_name} This configuration will display a "{deploy_name}(Azure)" option in the model list. The model used when making requests is gpt-3.5-turbo

limour-blog commented 2 months ago

写成下面这样就行

    environment:
      - CUSTOM_MODELS=-all,+gpt-3.5-turbo@openai,+gpt-4-turbo@openai,+gpt-4o@openai,+claude-3-haiku