CamemBERT: a Tasty French Language Model

Introduction

CamemBERT is a state-of-the-art language model for French based on the RoBERTa model.

It is now available on Hugging Face in 6 different versions with varying number of parameters, amount of pretraining data and pretraining data source domains.

For further information or requests, please go to Camembert Website

Pre-trained models

Model	#params	Arch.	Training data
`camembert-base`	110M	Base	OSCAR (138 GB of text)
`camembert/camembert-large`	335M	Large	CCNet (135 GB of text)
`camembert/camembert-base-ccnet`	110M	Base	CCNet (135 GB of text)
`camembert/camembert-base-wikipedia-4gb`	110M	Base	Wikipedia (4 GB of text)
`camembert/camembert-base-oscar-4gb`	110M	Base	Subsample of OSCAR (4 GB of text)
`camembert/camembert-base-ccnet-4gb`	110M	Base	Subsample of CCNet (4 GB of text)

How to use CamemBERT with HuggingFace

Load CamemBERT and its sub-word tokenizer :

from transformers import CamembertModel, CamembertTokenizer

# You can replace "camembert-base" with any other model from the table, e.g. "camembert/camembert-large".
tokenizer = CamembertTokenizer.from_pretrained("camembert/camembert-base-ccnet")
camembert = CamembertModel.from_pretrained("camembert/camembert-base-ccnet")

camembert.eval()  # disable dropout (or leave in train mode to finetune)

Filling masks using pipeline

from transformers import pipeline 

camembert_fill_mask  = pipeline("fill-mask", model="camembert/camembert-base-ccnet", tokenizer="camembert/camembert-base-ccnet")
results = camembert_fill_mask("Le camembert est <mask> :)")
# results
#[{'sequence': '<s> Le camembert est bon :)</s>', 'score': 0.14011502265930176, 'token': 305},
# {'sequence': '<s> Le camembert est délicieux :)</s>', 'score': 0.13929404318332672, 'token': 11661}, 
# {'sequence': '<s> Le camembert est excellent :)</s>', 'score': 0.07010319083929062, 'token': 3497}, 
# {'sequence': '<s> Le camembert est parfait :)</s>', 'score': 0.025885622948408127, 'token': 2528}, 
# {'sequence': '<s> Le camembert est top :)</s>', 'score': 0.025684962049126625, 'token': 2328}]

Extract contextual embedding features from Camembert output

import torch
# Tokenize in sub-words with SentencePiece
tokenized_sentence = tokenizer.tokenize("J'aime le camembert !")
# ['▁J', "'", 'aime', '▁le', '▁cam', 'ember', 't', '▁!'] 

# 1-hot encode and add special starting and end tokens 
encoded_sentence = tokenizer.encode(tokenized_sentence)
# [5, 133, 22, 1250, 16, 12034, 14324, 81, 76, 6]
# NB: Can be done in one step : tokenize.encode("J'aime le camembert !")

# Feed tokens to Camembert as a torch tensor (batch dim 1)
encoded_sentence = torch.tensor(encoded_sentence).unsqueeze(0)
embeddings, _ = camembert(encoded_sentence)
# embeddings.detach()
# embeddings.size torch.Size([1, 10, 768])
#tensor([[[ 0.0667, -0.2467,  0.0954,  ...,  0.2144,  0.0279,  0.3621],
#         [-0.0472,  0.4092, -0.6602,  ...,  0.2095,  0.1391, -0.0401],
#         [ 0.1911, -0.2347, -0.0811,  ...,  0.4306, -0.0639,  0.1821],
#         ...,

Extract contextual embedding features from all Camembert layers

from transformers import CamembertConfig
# (Need to reload the model with new config)
config = CamembertConfig.from_pretrained("camembert/camembert-base-ccnet", output_hidden_states=True)
camembert = CamembertModel.from_pretrained("camembert/camembert-base-ccnet", config=config)

embeddings, _, all_layer_embeddings = camembert(encoded_sentence)
#  all_layer_embeddings list of len(all_layer_embeddings) == 13 (input embedding layer + 12 self attention layers)
all_layer_embeddings[5]
# layer 5 contextual embedding : size torch.Size([1, 10, 768])
#tensor([[[ 0.0057, -0.1022,  0.0163,  ..., -0.0675, -0.0360,  0.1078],
#         [-0.1096, -0.3344, -0.0593,  ...,  0.1625, -0.0432, -0.1646],
#         [ 0.3751, -0.3829,  0.0844,  ...,  0.1067, -0.0330,  0.3334],
#         ...,

Authors

CamemBERT was trained and evaluated by Louis Martin*, Benjamin Muller*, Pedro Javier Ortiz Suárez*, Yoann Dupont, Laurent Romary, Éric Villemonte de la Clergerie, Djamé Seddah and Benoît Sagot.

Citation

If you use our work, please cite:

@inproceedings{martin2020camembert,
  title={CamemBERT: a Tasty French Language Model},
  author={Martin, Louis and Muller, Benjamin and Su{\'a}rez, Pedro Javier Ortiz and Dupont, Yoann and Romary, Laurent and de la Clergerie, {\'E}ric Villemonte and Seddah, Djam{\'e} and Sagot, Beno{\^\i}t},
  booktitle={Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},
  year={2020}
}

almanach
/

camembert-base-ccnet