Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

ViGPTQA-state-of-the-art LLMs for vietnamese question answering : system overview, core models training, and evaluations

Nguyen, Minh Thuan ; Tran, Khanh-Tung ; Nguyen, Vincent and Vu, Xuan-Son LU (2023) p.754-764
Abstract
Large language models (LLMs) and their applications in low-resource languages (such as in Vietnamese) are limited due to lack of training data and benchmarking datasets. This paper introduces a practical real-world implementation of a question answering system for Vietnamese, called ViGPTQA, leveraging the power of LLM. Since there is no effective LLM in Vietnamese to date, we also propose, evaluate, and open-source an instruction-tuned LLM for Vietnamese, named ViGPT. ViGPT demonstrates exceptional performances, especially on real-world scenarios. We curate a new set of benchmark datasets that encompass both AI and human-generated data, providing a comprehensive evaluation framework for Vietnamese LLMs. By achieving state-of-the-art... (More)
Large language models (LLMs) and their applications in low-resource languages (such as in Vietnamese) are limited due to lack of training data and benchmarking datasets. This paper introduces a practical real-world implementation of a question answering system for Vietnamese, called ViGPTQA, leveraging the power of LLM. Since there is no effective LLM in Vietnamese to date, we also propose, evaluate, and open-source an instruction-tuned LLM for Vietnamese, named ViGPT. ViGPT demonstrates exceptional performances, especially on real-world scenarios. We curate a new set of benchmark datasets that encompass both AI and human-generated data, providing a comprehensive evaluation framework for Vietnamese LLMs. By achieving state-of-the-art results and approaching other multilingual LLMs, our instruction-tuned LLM underscores the need for dedicated Vietnamese-specific LLMs. Our open-source model supports customized and privacy-fulfilled Vietnamese language processing systems. (Less)
Please use this url to cite or link to this publication:
author
; ; and
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
host publication
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing : industry Track - industry Track
editor
Wang, Mingxuan and Zitouni, Imed
pages
754 - 764
publisher
Association for Computational Linguistics
external identifiers
  • scopus:85180767795
DOI
10.18653/v1/2023.emnlp-industry.70
language
English
LU publication?
no
id
e9a2135b-41c5-4145-96e6-6c1b90022303
date added to LUP
2026-02-10 23:56:07
date last changed
2026-03-10 10:23:24
@inproceedings{e9a2135b-41c5-4145-96e6-6c1b90022303,
  abstract     = {{Large language models (LLMs) and their applications in low-resource languages (such as in Vietnamese) are limited due to lack of training data and benchmarking datasets. This paper introduces a practical real-world implementation of a question answering system for Vietnamese, called ViGPTQA, leveraging the power of LLM. Since there is no effective LLM in Vietnamese to date, we also propose, evaluate, and open-source an instruction-tuned LLM for Vietnamese, named ViGPT. ViGPT demonstrates exceptional performances, especially on real-world scenarios. We curate a new set of benchmark datasets that encompass both AI and human-generated data, providing a comprehensive evaluation framework for Vietnamese LLMs. By achieving state-of-the-art results and approaching other multilingual LLMs, our instruction-tuned LLM underscores the need for dedicated Vietnamese-specific LLMs. Our open-source model supports customized and privacy-fulfilled Vietnamese language processing systems.}},
  author       = {{Nguyen, Minh Thuan and Tran, Khanh-Tung and Nguyen, Vincent and Vu, Xuan-Son}},
  booktitle    = {{Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing : industry Track}},
  editor       = {{Wang, Mingxuan and Zitouni, Imed}},
  language     = {{eng}},
  pages        = {{754--764}},
  publisher    = {{Association for Computational Linguistics}},
  title        = {{ViGPTQA-state-of-the-art LLMs for vietnamese question answering : system overview, core models training, and evaluations}},
  url          = {{http://dx.doi.org/10.18653/v1/2023.emnlp-industry.70}},
  doi          = {{10.18653/v1/2023.emnlp-industry.70}},
  year         = {{2023}},
}