metadata

pipeline_tag: fill-mask
widget:
  - text: đậu xanh rau <mask>

ViSoBERT: A Pre-Trained Language Model for Vietnamese Social Media Text Processing (EMNLP 2023 - Main)

Disclaimer: The paper contains actual comments on social networks that might be construed as abusive, offensive, or obscene.

ViSoBERT is the state-of-the-art language model for Vietnamese social media tasks:

ViSoBERT is the first monolingual MLM (XLM-R architecture) built specifically for Vietnamese social media texts.
ViSoBERT outperforms previous monolingual, multilingual, and multilingual social media approaches, obtaining new state-of-the-art performances on four downstream Vietnamese social media tasks.

The general architecture and experimental results of ViSoBERT can be found in our paper:

@misc{nguyen2023visobert,
      title={ViSoBERT: A Pre-Trained Language Model for Vietnamese Social Media Text Processing}, 
      author={Quoc-Nam Nguyen and Thang Chau Phan and Duc-Vu Nguyen and Kiet Van Nguyen},
      year={2023},
      eprint={2310.11166},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Please CITE our paper when ViSoBERT is used to help produce published results or is incorporated into other software.

Installation

Install transformers and SentencePiece packages:

pip install transformers
pip install SentencePiece

Example usage

from transformers import AutoModel, AutoTokenizer
import torch

model= AutoModel.from_pretrained('uitnlp/visobert')
tokenizer = AutoTokenizer.from_pretrained('uitnlp/visobert')

encoding = tokenizer('dau xanh rau ma',return_tensors='pt')

with torch.no_grad():
  output = model(**encoding)