Lapsang 山正
Install
As we will import 7 million sentences from Tatoeba, it's better for you to use PostgreSQL.
This will install the Django development version, so that we can use the ignore_conflicts flag of bulk_create when importing the Tatoeba sentences.
python3 -m venv venv
. venv/bin/activate
pip install -r requirements.txt
./manage.py migrate
Download Tatoeba's sentences and links files, then:
./manage.py import --sentences /path/to/sentences.csv --links /path/to/links.csv # Should take 30 min
./manage.py runserver
Then you can access the routes http://localhost:8000/sentences/eng for example, or http://localhost:8000/sentence/1.