HerBERT: Efficiently Pretrained Transformer-based Language Model for Polish
arXiv:2105.01735
Abstract
BERT-based models are currently used for solving nearly all Natural Language Processing (NLP) tasks and most often achieve state-of-the-art results. Therefore, the NLP community conducts extensive research on understanding these models, but above all on designing effective and efficient training procedures. Several ablation studies investigating how to train BERT-like models have been carried out, but the vast majority of them concerned only the English language. A training procedure designed for English does not have to be universal and applicable to other especially typologically different languages. Therefore, this paper presents the first ablation study focused on Polish, which, unlike the isolating English language, is a fusional language. We design and thoroughly evaluate a pretraining procedure of transferring knowledge from multilingual to monolingual BERT-based models. In addition to multilingual model initialization, other factors that possibly influence pretraining are also explored, i.e. training objective, corpus size, BPE-Dropout, and pretraining length. Based on the proposed procedure, a Polish BERT-based language model -- HerBERT -- is trained. This model achieves state-of-the-art results on multiple downstream tasks.
arXiv commentPublished in Proceedings of the 8th Workshop on Balto-Slavic Natural Language Processing
References in corpus (8)
- Distributed Representations of Words and Phrases and their Compositionality
- Language Models are Few-Shot Learners
- Scaling Laws for Neural Language Models
- CamemBERT: a Tasty French Language Model
- Adaptation of Deep Bidirectional Multilingual Transformers for Russian Language
- Multilingual is not enough: BERT for Finnish
- FRAGE: Frequency-Agnostic Word Representation
- CLUE: A Chinese Language Understanding Evaluation Benchmark