

Introduction

django-ner-trainer is a set of Django models, management commands, and REST endpoints to help make it easier to train spaCy Named Entity Recognition models within Django.

N.B. This package is designed to provide the minimum backend infrastructure you need to train an NER model. It isn't meant to be a full-featured application like Prodigy. In other words, this plugin is bring-your-own-front-end.

Quickstart

Install Django NER Trainer:

pip install django-ner-trainer

Add it to your INSTALLED_APPS:

INSTALLED_APPS = (
    ...
    'ner_trainer',
    ...
)

Add Django NER Trainer's URL patterns:

from ner_trainer import urls as ner_trainer_urls


urlpatterns = [
    ...
    url(r'^', include(ner_trainer_urls)),
    ...
]

Features

• Management commands to download spaCy models, training a NER model, and packaging a NER model
• Models for storing phrases, entities, and tagged phrase entities
• REST Endpoints for performing CRUD actions on relevant models

Running Tests

Does the code actually work?

source <YOURVIRTUALENV>/bin/activate
(myenv) $ pip install tox
(myenv) $ tox