paper-with-me

홈 › Papers

BARThez: a Skilled Pretrained French Sequence-to-Sequence Model

2020-10-23 · EMNLP 2021 11 · Moussa Kamal Eddine, Antoine J. -P. Tixier, Michalis Vazirgiannis

Inductive transfer learning has taken the entire NLP field by storm, with models such as BERT and BART setting new state of the art on countless NLU tasks. However, most of the available models and research have been conducted for English. In this work, we introduce BARThez, the first large-scale pretrained seq2seq model for French. Being based on BART, BARThez is particularly well-suited for generative tasks. We evaluate BARThez on five discriminative tasks from the FLUE benchmark and two generative tasks from a novel summarization dataset, OrangeSum, that we created for this research. We show BARThez to be very competitive with state-of-the-art BERT-based French language models such as CamemBERT and FlauBERT. We also continue the pretraining of a multilingual BART on BARThez' corpus, and show our resulting model, mBARThez, to significantly boost BARThez' generative performance. Code, data and models are publicly available.

📄 PDF Abstract BibTeX arXiv:2010.12321

Code (5)

moussaKam/BARThez 공식 구현 pytorch
2024-MindSpore-1/Code2/tree/main/model-1/barthez mindspore
Tixierae/OrangeSum
huggingface/transformers pytorch
moussaKam/OrangeSum

Tasks

FLUENatural Language UnderstandingOrangeSumSelf-Supervised LearningText SummarizationTransfer Learning

Methods 이 논문이 사용한 방법론

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
mBARTHez BARThez is a self-supervised transfer learning model for the French language based on BART. Compared to existing…
Sigmoid Activation 설명 없음
Tanh Activation 설명 없음
LSTM An LSTM is a type of recurrent neural network that addresses the vanishing gradient problem in vanilla…
Seq2Seq Seq2Seq, or Sequence To Sequence, is a model used in sequence prediction tasks, such as language modelling and machine translation. The idea is to use one…
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Adam 설명 없음

Similar Papers 제목 키워드 기반

Effectiveness of French Language Models on Abstractive Dialogue Summarization Task

2022-07-17 · LREC 2022 6 · Yongxin Zhou, François Portet, Fabien Ringeval

Pre-trained language models have established the state-of-the-art on various natural language processing tasks, including dialogue summarization, which allows the reader to quickly access key information from long conver…

Abstractive Dialogue Summarization

Adaptation of Biomedical and Clinical Pretrained Models to French Long Documents: A Comparative Study

2024-02-26 · Adrien Bazoge, Emmanuel Morin, Beatrice Daille, Pierre-Antoine Gourraud

Recently, pretrained language models based on BERT have been introduced for the French biomedical domain. Although these models have achieved state-of-the-art results on biomedical and clinical NLP tasks, they are constr…

named-entity-recognitionNamed Entity Recognition

Using Multimodal and Language-Agnostic Sentence Embeddings for Abstractive Summarization

2026-03-09 · Chaimae Chellaf, Salima Mdhaffar, Yannick Estève, Stéphane Huet arxiv

Abstractive summarization aims to generate concise summaries by creating new sentences, allowing for flexible rephrasing. However, this approach can be vulnerable to inaccuracies, particularly `hallucinations' where the …

GAFSV-Net: A Vision Framework for Online Signature Verification

2026-04-30 · Himanshu Singhal, Suresh Sundaram arxiv

Online signature verification (OSV) requires distinguishing skilled forgeries from genuine samples under high intra-class variability and with very few enrollment samples. Existing deep learning methods operate directly …

Temporal Sequences

Low-resource Neural Machine Translation: Benchmarking State-of-the-art Transformer for Wolof<->French

2022-06-01 · LREC 2022 6 · Cheikh M. Bamba Dione, Alla Lo, Elhadji Mamadou Nguer, Sileye Ba

In this paper, we propose two neural machine translation (NMT) systems (French-to-Wolof and Wolof-to-French) based on sequence-to-sequence with attention and Transformer architectures. We trained our models on the parall…

BenchmarkingLow Resource Neural Machine TranslationLow-Resource Neural Machine TranslationMachine Translation+3