ERNIE 2.0: A Continual Pre-training Framework for Language Understanding

TL;DR

ERNIE 2.0 introduces a continual multi-task pre-training framework that incrementally learns lexical, syntactic, and semantic information, significantly improving performance across various language understanding benchmarks.

Contribution

It presents a novel continual pre-training approach that leverages multiple tasks to enhance language models beyond traditional single-task pre-training methods.

Findings

Outperforms BERT and XLNet on 16 language understanding tasks.

Achieves state-of-the-art results on GLUE benchmarks and Chinese tasks.

Demonstrates the effectiveness of multi-task continual learning in pre-training.

Abstract

Recently, pre-trained models have achieved state-of-the-art results in various language understanding tasks, which indicates that pre-training on large-scale corpora may play a crucial role in natural language processing. Current pre-training procedures usually focus on training the model with several simple tasks to grasp the co-occurrence of words or sentences. However, besides co-occurring, there exists other valuable lexical, syntactic and semantic information in training corpora, such as named entity, semantic closeness and discourse relations. In order to extract to the fullest extent, the lexical, syntactic and semantic information from training corpora, we propose a continual pre-training framework named ERNIE 2.0 which builds and learns incrementally pre-training tasks through constant multi-task learning. Experimental results demonstrate that ERNIE 2.0 outperforms BERT and XLNet on 16 tasks including English tasks on GLUE benchmarks and several common tasks in Chinese. The source codes and pre-trained models have been released at https://github.com/PaddlePaddle/ERNIE.