# Parallel corpus training

A model based on the ModernBERT architecture, trained on a 24GB parallel sentence corpus from OpenSubtitles with Polish language alignment using cross-tokenization technology.

This model is a fine-tuned Korean-to-English translation model based on Helsinki-NLP/opus-mt-ko-en, trained using the Korean Parallel Corpora dataset.

A BART-based Cantonese to Traditional Chinese translation model fine-tuned on Cantonese-Traditional Chinese parallel corpus

BigTranslate is a multilingual translation model enhanced based on LLaMA-13B, supporting translation tasks in over 100 languages.

A Russian text detoxification model trained on ruT5-base, used to convert toxic Russian text into neutral expressions.

Vietnamese-English machine translation model based on T5 architecture