HerBERT: Efficiently pretrained transformer-based language model for Polish

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F21%3A10440950" target="_blank" >RIV/00216208:11320/21:10440950 - isvavai.cz</a>

Result on the web

—

DOI - Digital Object Identifier

—

Result language

angličtina

Original language name

HerBERT: Efficiently pretrained transformer-based language model for Polish

Original language description

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.

Czech name

—

Czech description

—

Type

D - Article in proceedings

CEP classification

—

OECD FORD branch

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Project

—

Continuities

—

Publication year

2021

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection

Proceedings of the 8th BSNLP Workshop on Balto-Slavic Natural Language Processing, BSNLP 2021 - Co-located with the 16th European Chapter of the Association for Computational Linguistics, EACL 2021

ISBN

978-1-954085-14-5

ISSN

—

e-ISSN

—

Number of pages

10

Pages from-to

1-10

Publisher name

Association for Computational Linguistics

Place of publication

Stroudsburg

Event location

online

Event date

Apr 20, 2021

Type of event by nationality

WRD - Celosvětová akce

UT code for WoS article

—