Papers › CamemBERT: a Tasty French Language Model

CamemBERT: a Tasty French Language Model

10 Nov 2019ACL 2020 6arXiv:1911.03894archive 2025-07-28

Louis Martin, Benjamin Muller, Pedro Javier Ortiz Suárez, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah, Benoît Sagot

Pretrained language models are now ubiquitous in Natural Language Processing. Despite their success, most available models have either been trained on English data or on the concatenation of data in multiple languages. This makes practical use of such models --in all languages except English-- very limited. In this paper, we investigate the feasibility of training monolingual Transformer-based language models for other languages, taking French as an example and evaluating our language models on part-of-speech tagging, dependency parsing, named entity recognition and natural language inference tasks. We show that the use of web crawled data is preferable to the use of Wikipedia data. More surprisingly, we show that a relatively small web crawled dataset (4GB) leads to results that are as good as those obtained using larger datasets (130+GB). Our best performing model CamemBERT reaches or improves the state of the art in all four downstream tasks.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

huggingface/transformers officialmentioned in papermentioned on GitHubpytorch report
anaishoareau/french_preprocessing mentioned on GitHubpytorch report
bourrel/French-News-Clustering mentioned on GitHubtf report
pwc-1/Paper-8 mindspore report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Dependency ParsingLanguage ModelingLanguage ModellingNamed Entity RecognitionNamed Entity Recognition (NER)Natural Language InferencePart-Of-Speech Taggingmodel

Datasets

Introduced by this paper, per the archive.

French Wikipedia

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Dependency Parsing French GSD CamemBERT LAS 92.47 #1 of 2 Archive leaderboard report
Dependency Parsing French GSD CamemBERT UAS 94.82 #1 of 2 Archive leaderboard report
Dependency Parsing ParTUT CamemBERT LAS 92.9 #1 of 2 Archive leaderboard report
Dependency Parsing ParTUT CamemBERT UAS 95.21 #1 of 2 Archive leaderboard report
Dependency Parsing Sequoia Treebank CamemBERT LAS 94.39 #1 of 2 Archive leaderboard report
Dependency Parsing Sequoia Treebank CamemBERT UAS 95.56 #1 of 2 Archive leaderboard report
Dependency Parsing Spoken Corpus CamemBERT LAS 81.37 #1 of 2 Archive leaderboard report
Dependency Parsing Spoken Corpus CamemBERT UAS 86.05 #1 of 2 Archive leaderboard report
Named Entity Recognition (NER) French Treebank CamemBERT (subword masking) F1 87.93 #1 of 1 Archive leaderboard report
Named Entity Recognition (NER) French Treebank CamemBERT (subword masking) Precision 88.35 #1 of 1 Archive leaderboard report
Named Entity Recognition (NER) French Treebank CamemBERT (subword masking) Recall 87.46 #1 of 1 Archive leaderboard report
Natural Language Inference XNLI French CamemBERT (large) Accuracy 85.7 #1 of 6 Archive leaderboard report
Natural Language Inference XNLI French CamemBERT (base) Accuracy 81.2 #3 of 6 Archive leaderboard report
Part-Of-Speech Tagging French GSD CamemBERT UPOS 98.19 #1 of 1 Archive leaderboard report
Part-Of-Speech Tagging ParTUT CamemBERT UPOS 97.63 #1 of 1 Archive leaderboard report
Part-Of-Speech Tagging Sequoia Treebank CamemBERT UPOS 99.21 #1 of 1 Archive leaderboard report
Part-Of-Speech Tagging Spoken Corpus CamemBERT UPOS 96.68 #1 of 1 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections