ERNIE 3.0: Large-scale Knowledge Enhanced Pre-training for Language Understanding and Generation

Authors:
  • YSYu Sun
    Yu Sun
  • SWShuohuan Wang
    Shuohuan Wang
  • SFShikun Feng
    Shikun Feng
  • SDSiyu Ding
    Siyu Ding
  • CPChao Pang
    Chao Pang
  • JSJunyuan Shang
    Junyuan Shang
  • JLJiaxiang Liu
    Jiaxiang Liu
  • XCXuyi Chen
    Xuyi Chen
  • YZYanbin Zhao
    Yanbin Zhao
  • YLYuxiang Lu
    Yuxiang Lu
  • WLWeixin Liu
    Weixin Liu
  • ZWZhihua Wu
    Zhihua Wu

Abstract
Pre-trained models have achieved state-of-the-art results in various Natural Language Processing (NLP) tasks. Recent works such as T5 and GPT-3 have shown that scaling up pre-trained language models can improve their generalization abilities. Particularly, the GPT-3 model with 175 billion parameters shows its strong task-agnostic zero-shot/few-shot learning capabilities. Despite their success, these large-scale models are trained on plain texts without introducing knowledge such as linguistic knowledge and world knowledge. In addition, most large-scale models are trained in an auto-regressive way. As a result, this kind of traditional fine-tuning approach demonstrates relatively weak performance when solving downstream language understanding tasks. In order to solve the above problems, we propose a unified framework named ERNIE 3.0 for pre-training large-scale knowledge enhanced models. It fuses auto-regressive network and auto-encoding network, so that the trained model can be easily tailored for both natural language understanding and generation tasks with zero-shot learning, few-shot learning or fine-tuning. We trained the model with 10 billion parameters on a 4TB corpus consisting of plain texts and a large-scale knowledge graph. Empirical results show that the model outperforms the state-of-the-art models on 54 Chinese NLP tasks, and its English version achieves the first place on the SuperGLUE benchmark (July 3, 2021), surpassing the human performance by +0.8% (90.6% vs. 89.8%).