Renaissance Labs

Atlas / Europe

Institute · Paris, France

ALMAnaCHInria

Inria Paris research team behind the CamemBERT family of open French language models and work in computational humanities.

NLPFrenchlanguage modelsdigital humanities

Paris, France

Type
Institute
Based in
Paris, France
Key papers
2
Featured projects
4

About

ALMAnaCH is an Inria team in Paris working on natural language processing and digital humanities. Its members include Benoît Sagot and Djamé Seddah, who are authors of the original CamemBERT paper alongside colleagues from Inria and Sorbonne Université.

CamemBERT, a RoBERTa-based French model released under an MIT licence in 2019, became a widely adopted reference. CamemBERT 2.0 followed in 2024, and CamemBERTa, by Wissam Antoun, Sagot and Seddah, explored data-efficient training. The team's Hugging Face organization lists 92 models and 43 datasets, including biomedical encoders and Gaperon sparse autoencoders.

The models serve researchers, companies and public bodies that need open French and biomedical text processing.

Mission and ambitions

The team builds and shares language models and data for French and historical or specialised text. It treats open release as part of academic practice.

Known for

  1. 01

    CamemBERT French language model

  2. 02

    CamemBERT 2.0 and CamemBERTa

  3. 03

    French biomedical encoders

  4. 04

    Digital humanities NLP

Milestones

  1. 2019

    CamemBERT

    Martin et al. released the reference open French language model (arXiv 1911.03894).

  2. 2023

    CamemBERTa

    Antoun, Sagot and Seddah published a data-efficient French language model.

  3. 2024

    CamemBERT 2.0

    Updated French encoders with a refreshed tokenizer and data were released (arXiv 2411.08868).

  4. 2025

    Biomedical models and data

    Published French biomedical encoders and datasets such as Biomed-Enriched.

Open source

Hugging Facehuggingface.co/almanach

Research papers

Notable releases

Openness

Models and datasets are public on Hugging Face under the almanach organisation; CamemBERT checkpoints have very high download counts.

People

Listed from the lab's own pages and public profiles.

Contact and community

Contact through the Inria team page.

Sources

Compiled from the lab's public pages, Hugging Face and GitHub. Last checked 2026-11-17. Something wrong or out of date? Tell us on Discord.

Labs nearby