tristan0x/ml-pic-annotate repository overview

⁠Picture Annotation Web Service for Machine Learning

⁠How to build the Docker image?

1. Install and start Docker

2. Run the following command:

   docker build -t ml-pic-annotate .
   

   Copy

⁠How to start a Docker container?

Build the Docker as described above. Run the command:

docker run --rm -ti -p 8080:8080 ml-pic-annotate

You can then open your browser at http://localhost:8080⁠

⁠uWSGI appplication

To start it:

uwsgi -H .env -s /tmp/ml-pic-annotate.sock \
  --env toto_proxy=toto \
  -w main:app \
  --daemonize $PWD/uwsgi.log \
  --logger file:$PWD/ml-pic-annotate-uwsgi.log \
  --pidfile $PWD/ml-pic-annotate-uwsgi.pid

Then to stop it:

uwsgi --stop $PWD/ml-pic-annotate-uwsgi.pid

⁠Development environment

⁠Start Flask application

Set up environment

virtualenv .env
.env/bin/activate
pip install -r requirements.txt

Start Flask application: python main.py

⁠Contributing

⁠Unit-tests

Before committing, ensure that Python unit-tests pass:

python tests/test_api.py

⁠Swagger

If you modify API, update swagger.yaml accordingly. You can advised to use the swagger online editor⁠

Please use scripts/swagger-check script to ensure the file is valid.

⁠Authors

Created by Francesco Casalegno, packaged by Tristan Carel