Skip to content

Latest commit

 

History

History
 
 

VECO

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
NLG
NLG
 
 
NLU
NLU
 
 
README.md
README.md
 
 

README.md

VECO: Variable Encoder-decoder Pre-training for Cross-lingual Understanding and Generation

Paper Link

News

  • May, 2021: VECO was accepted by ACL 2021.
  • Mar, 2021: VECO ranked first at the XTREME leaderboard.

Introduction

VECO is a variable encoder-decoder (VECO) model targets at providing pre-trained model initialization for both the encoder-only and encoder-decoder Transformer with the most streamlined parameters. . As a result, VECO delivers new state-of-the-art results on various cross-lingual understanding tasks of the XTREME benchmark covering text classification, sequence labeling, question answering, and sentence retrieval. For generation tasks, VECO also outperforms all existing cross-lingual models and state-of-the-art Transformer variants on WMT14 English-to-German and English-to-French translation datasets, with gains of up to 1∼2 BLEU.

Pre-trained models

Pre-trained models for Multilingual NLU tasks

Model Description # Params Download
VECO_large VECO model with 24 encoder layers trained on 50 languages' monolingual and bilingual corpus 550M veco-large.tar.gz

Pre-trained models for Multilingual NLG tasks

Model Description # Params Download
VECO_large VECO model with 24 encoder and decoder layers trained on 50 languages' monolingual and bilingual corpus 660M veco-large.tar.gz

Results

Results of Multilingual NLU tasks

XTREME is one of the most representative massively multilingual benchmark.

The results of XNLI, XQuAD and Tatoeba task can be reproduced using the default hyperparameters listed in .sh file.

Model XNLI
(Acc)		 XQuAD
(F1/EM)		 Tatoeba
(Acc)
VECO_large 79.9 77.5/61.9 75.1

Results of Multilingual NLG tasks

Results on custom machine translation datasets.

Model WMT14 En-Fr
tok/detok-BLEU		 WMT14 En-De
tok/detok-BLEU
XLM-R (24 encoder + 6 decoder) 43.7/41.1 30.8/29.9
mBART (12 encoder + 12 decoder) 43.2/41.0 30.0/29.1
VECO_large (24 encoder + 6 decoder) 44.4/42.0 31.5/30.5

Finetuning

Citation

@article{Luo2020VECO
  author    = {Fuli Luo and Wei Wang and Jiahao Liu and Yijia Liu and Bin Bi and Songfang Huang and Fei Huang and Luo Si},
  title     = {{VECO:} Variable Encoder-decoder Pre-training for Cross-lingual Understanding and Generation},
  journal   = {CoRR},
  volume    = {abs/2010.16046},
  year      = {2020},
  url       = {https://arxiv.org/abs/2010.16046},
  archivePrefix = {arXiv},
  eprint    = {2010.16046},
}