Table of Contents
Question Answering AI project using ChatGPT to respond to questions, Vector Stores to efficiently handle, retrieve information from the context files and Discord as interface.
- Disnake
- OpenAi
To be able to respond to the questions, I integrated the openAI API and used a vector store to enhance the speed of data retrieval for the custom files and reduce cost. I'm using Discord as interface for the project, therefore I used Disnake to interact with the Discord API.
- Docker
- Discord Bot Token
- Creat a free discord bot Discord developers portal to get a bot token
- Invite you new bot with admin permissions in your discord server
- Clone the repo
git clone https://github.com/vannsoko/DocumentQA.git
- Renaming .env.exemple to .env
- Enter BOT_TOKEN and START_CHANNEL_ID in
.envBOT_TOKEN="REPLACE THIS WITH THE BOT TOKEN FROM THE DISCORD DEVELOPERS PORTAL" START_CHANNEL_ID="START_CHANNEL ID WHERE THE BOT WILL BE INITIALIZED" OPENAI_API_KEY="" IN_DEV=true
- Find an issue or create one your self. If you have create the issue please make sure that it is relevant to the project.
- Create your Feature Branch (
git checkout -b feature/AmazingFeature). Please use the name of the issue in the branch name. or reference the issue id in the name. - Do your changes and commit them (
git commit -m 'Add some AmazingFeature') - Push to the Branch (
git push origin feature/AmazingFeature) - Open a Pull Request and wait for it to be merged