Abstractive Text Summarization Using the BRIO Training Paradigm

TL;DR

This paper introduces the BRIO training paradigm for abstractive summarization, enhancing model performance and control, especially for Vietnamese, by fine-tuning pre-trained language models with a non-deterministic approach.

Contribution

It presents a novel training paradigm, BRIO, that improves abstractive summarization by reducing dependence on reference summaries and demonstrates its effectiveness on Vietnamese and English datasets.

Findings

Models trained with BRIO outperform existing methods.

Significant improvements in Vietnamese summarization.

Effective on basic hardware.

Abstract

Summary sentences produced by abstractive summarization models may be coherent and comprehensive, but they lack control and rely heavily on reference summaries. The BRIO training paradigm assumes a non-deterministic distribution to reduce the model's dependence on reference summaries, and improve model performance during inference. This paper presents a straightforward but effective technique to improve abstractive summaries by fine-tuning pre-trained language models, and training them with the BRIO paradigm. We build a text summarization dataset for Vietnamese, called VieSum. We perform experiments with abstractive summarization models trained with the BRIO paradigm on the CNNDM and the VieSum datasets. The results show that the models, trained on basic hardware, outperform all existing abstractive summarization models, especially for Vietnamese.