Skip to main content

A graph dependency parser to tinker with

Project description

HOPS, an honest parser of sentences

Latest PyPI version Build Status Code style: black

It ain't much but it's honest work.

This is a graph-based dependency parser inspired by Dozat and Manning (2017)'s biaffine graph parser. Contrary to Dozat, the parser performs its own tagging and can use several lexers such as FastText, BERT and others. It has been originally designed within the FlauBERT initiative.

The parser comes with pretrained models ready for parsing French, but it might be trained for other languages without difficulties.

See the documentation for more information.


If you use this parser for your scientific publication, or if you find the resources in this repository useful, please cite the following paper

    title = {{Analyse en dépendances du français avec des plongements contextualisés}},
    author = {Grobol, Loïc and Crabbé, Benoît},
    url = {},
    booktitle = {{Actes de la 28ème Conférence sur le Traitement Automatique des Langues Naturelles}},
    eventtitle = {{TALN-RÉCITAL 2021}},
    venue = {Lille, France},
    pdf = {},
    hal_id = {hal-03223424},
    hal_version = {v1},


If you want a development install (so you can modify the code locally and directly run it), you can install it in editable mode with the tests extras after cloning the repository

git clone
cd hopsparser
pip install -e ".[tests,traintools]"

In that case, you can run the smoke tests with tox to ensure that everything works on your end.

Note that using the editable mode requires pip >= 21.3.1.


This software is released under the MIT Licence, with some files released under compatible free licences, see for the details.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

hopsparser-0.7.1.tar.gz (51.9 kB view hashes)

Uploaded source

Built Distribution

hopsparser-0.7.1-py3-none-any.whl (48.4 kB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page