NEZHA: Neural Contextualized Representation for Chinese Language Understanding

By Junqiu Wei, Xiaozhe Ren, Xiaoguang Li, WenYong Huang, Yi Liao, Yasheng Wang, Jiashu Lin, Xin Jiang, Xiao Chen, Qun Liu. 2019

128 citations recorded by the source corpus.

The pre-trained language models have achieved great successes in various natural language understanding (NLU) tasks due to its capacity to capture the deep contextualized information in text by pre-training on large-scale corpora. In this technical report, we present our practice of pre-training language models named NEZHA (NEural contextualiZed representation for CHinese lAnguage understanding) on Chinese corpora and finetuning for the Chinese NLU tasks. The current version of NEZHA is based on BERT with a collection of proven improvements, which include Functional Relative Positional Encoding as an effective positional encoding scheme, Whole Word Masking strategy, Mixed Precision Training and the LAMB Optimizer in training the models. The experimental results show that NEZHA achieves the state-of-the-art performances when finetuned on several representative Chinese tasks, including named entity recognition (People's Daily NER), sentence matching (LCQMC), Chinese sentiment classification (ChnSenti) and natural language inference (XNLI).

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Large Language Models, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Large Language Models

In Natural Language Processing

Citations in the corpus

RoFormer: Enhanced Transformer with Rotary Position Embedding (2021, methodology): “BERT[8], WoBERT[36], and NEZHA[39], we tabulate their tokenization level and position embedding information in table 3.”
Pre-Trained Models: Past, Present and Future (2021, background): “Similar ideas have also been explored in ERNIE (Sun et al., 2019c,d) (where a whole entity is masked), NEZHA (Wei et al., 2019), and Whole Word Masking (Cui et al., 2019).”
Pre-trained models for natural language processing: A survey (2020, background): “Language-Specific ERNIE(Baidu) [170], BERT-wwm-Chinese [29], NEZHA [198], ZEN [37] BERTje [33], CamemBERT [125], FlauBERT [95], RobBERT [35]”
CPM: A Large-scale Generative Chinese Pre-trained Language Model (2020, neutral): “Although there are some previous works providing powerful Chinese pre-trained language models (Cui et al., 2020; Xu et al., 2020; Wei et al., 2019; Sun et al., 2019; Cui et al., 2019a), their capabilities are limited due to the model size.”
Revisiting Pre-Trained Models for Chinese Natural Language Processing (2020, methodology): “In Chinese condition, WordPiece tokenizer no longer split the word into small fragments, as Chinese characters are not formed by alphabet-like symbols.”
Censorship of Online Encyclopedias: Implications for NLP Models (2021, unknown): “Perhaps because of the size difference between the two corpuses, increasingly researchers developing cutting edgeChinese langauge NLP models are drawing on the Baidu Baike corpus [38, 43].”
FewCLUE: A Chinese Few-shot Learning Evaluation Benchmark (2021, methodology): “GPT model is impemented using NEZHA-Gen (Wei et al., 2019) 3 .”
Training Multilingual Pre-trained Language Model with Byte-level Subwords (2021, methodology): “In the experiment, we adopted the architecture of NEZHA [5] as the underlying pre-trained language model and the results show that NEZHA trained with byte-level subwords consistently outperforms Google multilingual BERT and vanilla NEZHA by a notable margin in several multilingual NLU tasks.”