ChatGPTNextWeb/NextChat

[Feature Request] Support for Github Models

Open

#5,637 建立於 2024年10月11日

在 GitHub 查看
 (0 留言) (0 反應) (0 負責人)TypeScript (87,992 star) (59,717 fork)batch import
enhancementhelp wanted

描述

🥰 Feature Description

I would love to use https://github.com/marketplace/models/

Here is where to get the access key: https://docs.github.com/en/authentication/keeping-your-account-and-data-secure/managing-your-personal-access-tokens

The api endpoint is https://models.inference.ai.azure.com

🧐 Proposed Solution

It seems like its the ordinary openai api but with different api url

📝 Additional Information

n/a

貢獻者指南

技術棧
typescript
領域
backend
議題類型
feature
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
2
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
1-3 hours
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
fresh
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
clear
前置要求
Familiarity with NextChat's provider architectureUnderstanding of OpenAI API compatibility
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
75
研究方向
Examine how existing model providers (e.g., OpenAI, Azure) are configured in the codebase, likely in a providers directory or configuration file. Add a new provider entry with base URL 'https://models.inference.ai.azure.com' and authentication using a GitHub personal access token. Refer to GitHub Models documentation for endpoint details and ensure compatibility with the existing API call patterns.