Towards Lithuanian grammatical error correction

TL;DR

This paper develops a Lithuanian grammatical error correction model using transformer architectures, achieving high accuracy and providing open-source code for further use and development.

Contribution

It introduces the first transformer-based grammatical error correction model for Lithuanian, comparing subword and byte-level approaches.

Findings

Achieved F₀.5=0.92 on Lithuanian error correction

Compared subword and byte-level transformer models

Provided open-source code for the model

Abstract

Everyone wants to write beautiful and correct text, yet the lack of language skills, experience, or hasty typing can result in errors. By employing the recent advances in transformer architectures, we construct a grammatical error correction model for Lithuanian, the language rich in archaic features. We compare subword and byte-level approaches and share our best trained model, achieving F$_{0.5}$=0.92, and accompanying code, in an online open-source repository.