Just send Dogbi a message with a photo of a dog or a person and it will analyze the image and respond with an estimate of similarity.
- Fill out
const_empty.py
in telegram_bot directory and rename it toconst.py
- Generate a legit secret key for
settings_without_key.py
(addSECRET_KEY = 'YOUR KEY'
) and rename it tosettings.py
- Copy the static files into AI directory (graph, breeds.csv, russian_breeds.csv should be placed in
root/AI
directory, while static images should be placed inroot/AI/static/media/
directory) - Build a docker image using
docker-compose build
and launch it withdocker-compose up
- Forward the
8005
and8006
ports to a default80
port of your server - That's it! Dogbi is now running in a Docker on your server!