Latest Progress 🎉

[March 2024] Add gemma-7b and qwen-7b models（based on Ollama）
[February 2024] Add mistral-7b model (based on Ollama)
[February 2024] Add gemini-pro model (based on Open API)
[January 2024] refactor the config-template.yaml to control the backend and the frontend settings at the same time, click to find more introduction about the config-template.yaml
[January 2024] Add internlm2-chat-7b model (based on LMDeploy)
[January 2024] Released version v0.0.1, officially open source！

Introduction

What is OpenAOE?

AOE, an acronym from DOTA2 for Area Of Effect, denotes an ability that can affect a group of targets within a certain area. Here, AOE in AI implies that user can obtain parallel outputs from multiple LLMs with one single prompt at the same time.

What problem does OpenAOE want to solve?

Currently, there are many open-source frameworks based on the ChatGPT for chat, but the LGC(LLM Group Chat) framework is still not coming yet.

The emergence of OpenAOE fills this gap: OpenAOE can help LLM researchers, evaluators, engineering developers, and even non-professionals to quickly access the market's well-known commercial and open-source LLMs, providing both single model serial response mode and multi-models parallel response mode.

What can you get from OpenAOE?

OpenAOE can:

return one or more LLMs' answers at the same time by a single prompt.
provide access to commercial LLM APIs, with default support for gpt3.5, gpt4, Google Palm, Minimax, Claude, Spark, etc., and also support user-defined access to other large model APIs. (API keys need to be prepared in advanced)
provide access to open-source LLM APIs. ( We recommend to use LMDeploy to deploy with one click)
provide backend APIs and a WEB-UI to meet the needs of different requirements.

Quick Run

Tip

Require python >= 3.9

Run by source code

Install

clone this project

git clone https://github.com/internlm/OpenAOE

build the frontend project when the frontend codes are changed (install node.js and npm first)

cd OpenAOE/openaoe/frontend
npm install
npm run build

Start

cd OpenAOE # this OpenAOE is the clone directory
pip install -r openaoe/backend/requirements.txt
# add your api key in config first
python -u -m openaoe.main -f openaoe/backend/config/config-template.yaml

Tip

/path/to/your/config-template.yaml is a configuration file loaded by OpenAOE at startup, which contains the relevant configuration information for the LLMs, including: API URLs, AKSKs, Tokens, etc. A template configuration yaml file can be found in openaoe/backend/config/config-template.yaml. Note that, this config-template.yaml DOES NOT contain any API access data, you should add them by yourself.

Tech Stack

The technology stack we use includes:

Backend framework based on python + fastapi;
Frontend framework based on typescript + Sealion-Client (encapsulated based on React) + Sealion-UI.
Build tools:
1. conda: quickly create a virtual python env to install necessary packages
2. npm: build the frontend project

Tip

The build tools can be installed quickly by pip install -U sealion-cli

Organization of the Repo

Frontend codes are in openaoe/frontend
Backend codes are in openaoe/backend
Project entry-point is openaoe/main.py

How to add a new model

Frontend

Add new model info like name, avatar, provider, etc in openaoe/frontend/src/config/model-config.ts
Add a new model basic API request payload configuration in openaoe/frontend/src/config/api-config.ts
Modify your new model's payload specifically in openaoe/frontend/src/services/fetch.ts, you may need to change the payload structure and handle corner cases according to your model's API definition.

PaParaZz1/OpenAOE