Atanet is a web application, where everybody can freely write posts and vote for posts others have already written. Those posts can be sorted with a variety of different filters. What makes Atanet special is, that every post will be given a class classified by a RNN implemented with TensorFlow. The technologies used in this project are ASP.NET Core, Flask, TensorFlow and Angular.

docker network create atanet-network

docker volume create atanet-db-volume

docker volume create atanet-py-volume

Generate a client id in your Google Cloud Console.

Add .env file and replace client_id with generated Google Client ID. Also update the clientId in app/src/assets/Development.config.json, app/src/assets/Production.config.json and app/src/assets/.config.json.

ASPNETCORE_ENVIRONMENT=Development
GOOGLE_CLIENT_ID=client_id
SCOREBOARD_ENTRIES=5

MYSQL_ROOT_PASSWORD=password
MYSQL_USER=atanet
MYSQL_PASSWORD=password
MYSQL_DATABASE=Atanet
DATABASE_PORT=3306
# either db or ip
DATABASE_HOST=db
MAX_IMAGE_INDEX=130
SENTIMENT_HOST=py
SENTIMENT_PORT=9003
ASPNETCORE_URLS=http://localhost:9000

Create certifictes for nginx reverse proxy: openssl req -x509 -newkey rsa:4096 -nodes -out ./nginx/config/cert.pem -keyout ./nginx/config/key.pem -days 365

docker-compose up --build

mysql -u atanet -h 172.19.0.1 -P 9001 -D Atanet -p

• cd py && pipenv shell
• Run python -m pip uninstall --yes tensorflow-gpu && python -m pip install tensorflow
• Revert with python -m pip uninstall --yes tensorflow && python -m pip install tensorflow-gpu