AraELECTRA: Pre-Training Text Discriminators for Arabic Language Understanding

TL;DR

AraELECTRA introduces a discriminator-based pretraining approach for Arabic NLP, outperforming existing models in various tasks with smaller size by using replaced token detection.

Contribution

The paper presents AraELECTRA, the first Arabic ELECTRA model trained with replaced token detection, improving performance over masked language models.

Findings

Outperforms current Arabic language models on multiple NLP tasks.

Achieves better results with smaller model size.

Demonstrates effectiveness of ELECTRA-style training for Arabic.

Abstract

Advances in English language representation enabled a more sample-efficient pre-training task by Efficiently Learning an Encoder that Classifies Token Replacements Accurately (ELECTRA). Which, instead of training a model to recover masked tokens, it trains a discriminator model to distinguish true input tokens from corrupted tokens that were replaced by a generator network. On the other hand, current Arabic language representation approaches rely only on pretraining via masked language modeling. In this paper, we develop an Arabic language representation model, which we name AraELECTRA. Our model is pretrained using the replaced token detection objective on large Arabic text corpora. We evaluate our model on multiple Arabic NLP tasks, including reading comprehension, sentiment analysis, and named-entity recognition and we show that AraELECTRA outperforms current state-of-the-art Arabic language representation models, given the same pretraining data and with even a smaller model size.