zomi-tedim-ai/zomi-tedim-bible-corpus

Zomi Tedim–Burmese Parallel Corpus (Bible Verses)

Dataset Summary

A parallel text corpus of Bible verses in Tedim (Zomi) and Burmese, extracted from USX source files. This dataset is intended as a foundational resource for Tedim-language NLP research, including machine translation, language modeling, and text generation for this low-resource language.

Dataset Structure

Each record contains the following fields:

Data is provided in JSON and JSONL formats.

Data Source

Verses were extracted from 20 USX Bible book files:

1CH, 1CO, 1JN, 1KI, 1PE, 1SA, 1TH, 1TI, 2CH, 2CO, 2JN, 2KI, 2PE, 2SA, 2TH, 2TI, 3JN, ACT, AMO, COL

This yielded 7,289 aligned verse pairs.

Intended Use

• —Low-resource language modeling for Tedim (Zomi)
• —Machine translation (Tedim ↔ Burmese ↔ English)
• —Linguistic and NLP research on Tedim/Chin languages
• —Foundation for future Tedim datasets covering non-religious domains

Limitations

• —Vocabulary and register are limited to Biblical text and may not reflect everyday spoken or written Tedim.
• —Coverage is currently limited to 20 of the 66 Bible books; more books will be added over time.
• —English alignment may be incomplete for some verses.

Roadmap

This dataset is part of an ongoing effort to build broader Tedim language resources, in collaboration with the zolai-ai project.

Creator

Thang Deih Piang — Zomi GPT AI creator

Citation

If you use this dataset, please cite it as:

Zomi Tedim–Burmese Parallel Bible Corpus

Contact

Maintained by Thang Deih Piang. For collaboration inquiries, please open a discussion on this dataset's Community tab.