Telegram WaifuGPT

An AI companion and assistant with a cute anime-girl personality!

Environment Variables

Before building and running your project, create a .env file that is going to be used when you run the start script, also create e .env.local file that will be used when you run the dev script, and add the following variables:

You need an OpenAI account to use the project

OPENAI_API_KEY=<your_api_key>
OPENAI_API_ORG=<your_org_key>
OPENAI_API_MODEL=<gpt_model> # eg: gpt-3.5-turbo
TELEGRAM_BOT_KEY=<your_bot_key>
POSTGRES_URI=http://localhost
POSTGRES_PORT=5432
POSTGRES_DATABASE=<your_postgres_database>
POSTGRES_USERNAME=<your_postgres_username>
POSTGRES_PASSWORD=<your_postgres_password>
SERVER_HOST=http://localhost
SERVER_PORT=3000
REDIS_HOST=redis # Set either the name of the service on docker-compose.yml or the redis url 
REDIS_PORT=6379
REDIS_PASSWORD=<your_redis_password>
#DEBUG="grammy*" # To debug grammy bot

Building

To create a production version of your app:

npm run build

You can run the build with npm run start.

You can also use the shortcut npm run dev

Docker Compose

In order to make deploys easier, and also add other services, like Redis, you can run the project with docker-compose:

Build the docker-compose:

docker-compose build

Run docker-compose detached from your console:

docker-compose up -d

Docker

To run the project using only Docker:

Build the docker image

docker build . -t telegram-waifu-gpt

Run the docker detached from your console with the flag -d

docker run -d -p 3000:3000 telegram-waifu-gpt 

You can also add a name to the docker using --name <my-container-name>