Spaces:
Runtime error
title: TrustworthyAI
emoji: 🚀
colorFrom: blue
colorTo: green
sdk: docker
sdk_version: '1.0'
app_file: backend/app.py
pinned: false
TrustworthyAI
Local Dev Env Setup
frontend npm run start
backend python app.py
Setup in a new EC2 instance
Install nginx
sudo apt install nginx
Install apps
sudo apt install certbot python3-certbot-nginx ranger docker.io
Configure nginx
sudo certbot --nginx -d guardai.io -d www.guardai.io
Add this to the top of the nginx.conf on the EC2 instance
limit_req_zone $binary_remote_addr zone=mylimit:10m rate=1r/s;
Add proxy_pass` to both 80 and 443 blocks
location / {
# Rate limits backend API calls
limit_req zone=mylimit;
proxy_pass http://127.0.0.1:3000/;
}
See an example of final nginx configuration in/deployment/default.
Finally, restart nginx
sudo systemctl restart nginx
Build and run docker
Build:
docker build --pull --rm -f Dockerfile -t trustworthy-ai .
To run the docker image locally:
docker run --rm -p 3000:80/tcp trustworthy-ai
To push docker image to the repository:
docker tag trustworthy-ai:latest paulcccccch/trustworthy-ai:latest
docker push paulcccccch/trustworthy-ai:latest
To pull docker:
docker login
docker run -p 3000:80 paulcccccch/trustworthy-ai:latest
Clean up:
sudo docker stop $(sudo docker ps -aq)
sudo docker rm $(sudo docker ps -aq)
sudo docker rmi $(sudo docker images -q)
Environment Setting:
Frontend:
a. Download "Node.js". https://nodejs.org/en/download
b. Once you have downloaded it successfully, run this code then you should be able to see the version number.
node --version
- c. Go to the 'frontend' directory and run this code and you will see a file called "node modules" in the frontend file.
npm install
- d. Run this code and a website should pop up.
npm run start
Backend:
a. Download "Miniconda". https://docs.conda.io/projects/miniconda/en/latest/miniconda-install.html
b. Once you have downloaded it successfully, run this code then you should be able to see the version number.
conda --version
c. Go to the 'backend' directory.
d. Create a virtual environment using this code:
conda create --name 'name_of_environment'
- e. Activate the environment you just created using this code:
conda activate 'name_of_environment'
You may check the existing environment using this code:
conda env list
- f. In the environment you just activated, install all the modules in requirements.txt
pip install -r requirements.txt
You can use this code to check what modules you already have in your environment:
conda list
- g. When all the modules are successfully downloaded, you should be able to run the whole project. Run this code to see if it works. (Be sure you are in the correct environment)
python app.py