ViT5: Pretrained Text-to-Text Transformer for Vietnamese Language Generation

TL;DR

ViT5 is a pretrained Transformer model tailored for Vietnamese language tasks, achieving state-of-the-art results in text summarization and competitive performance in named entity recognition, highlighting the importance of context length during training.

Contribution

This work introduces ViT5, a Vietnamese-specific pretrained encoder-decoder Transformer model, and demonstrates its effectiveness on key language generation tasks with extensive experiments.

Findings

ViT5 outperforms existing models in Vietnamese text summarization.

ViT5 achieves competitive results in Vietnamese Named Entity Recognition.

Context length during pretraining significantly impacts downstream performance.

Abstract

We present ViT5, a pretrained Transformer-based encoder-decoder model for the Vietnamese language. With T5-style self-supervised pretraining, ViT5 is trained on a large corpus of high-quality and diverse Vietnamese texts. We benchmark ViT5 on two downstream text generation tasks, Abstractive Text Summarization and Named Entity Recognition. Although Abstractive Text Summarization has been widely studied for the English language thanks to its rich and large source of data, there has been minimal research into the same task in Vietnamese, a much lower resource language. In this work, we perform exhaustive experiments on both Vietnamese Abstractive Summarization and Named Entity Recognition, validating the performance of ViT5 against many other pretrained Transformer-based encoder-decoder models. Our experiments show that ViT5 significantly outperforms existing models and achieves state-of-the-art results on Vietnamese Text Summarization. On the task of Named Entity Recognition, ViT5 is competitive against previous best results from pretrained encoder-based Transformer models. Further analysis shows the importance of context length during the self-supervised pretraining on downstream performance across different settings.