CoolFace
Datasetpublic

seeksharp/tarok-nlp-corpus

Tarok New Testament Corpus Curated by SeekSharp Labs as part of a broader effort to build open NLP resources for Tarok (also known as Yergam, ISO 639-3: yer), a Plateau language spoken mainly in Langtang-North, Langtang-South, Wase, Mikang and Kanke LGAs of Plateau State, Nigeria. Source and License This dataset is derived from the Tarok New Testament translation published at ebible.org/pdf/yer, licensed under Creative Commons Attribution-ShareAlike 4.0… See the full description on the dataset page: https://huggingface.co/datasets/seeksharp/tarok-nlp-corpus.

sourceHugging Facecc-by-sa-4.0updated 2d agoView on Hugging Face
1likes63downloads
Dataset Card

Tarok New Testament Corpus

Curated by SeekSharp Labs as part of a broader effort to build open NLP resources for Tarok (also known as Yergam, ISO 639-3: yer), a Plateau language spoken mainly in Langtang-North, Langtang-South, Wase, Mikang and Kanke LGAs of Plateau State, Nigeria.

Dataset Description

  • —Curated by: SeekSharp Labs (Founder: Pirfa)
  • —Language: Tarok / Yergam (yer)
  • —Source text: Tarok New Testament, eBible.org
  • —Content: 3,699 unique sentences (verses), split 90/10 into train/test
  • —Format: JSON Lines (.jsonl), one sentence per line

Source and License

This dataset is derived from the Tarok New Testament translation published at ebible.org/pdf/yer, licensed under Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0).

Original copyright: © 2024 Luke Initiative for Scripture Translation, contributor: The Seed Company.

In line with the CC BY-SA 4.0 license:

  • —This derived dataset is released under the same CC BY-SA 4.0 license.
  • —The original text has been extracted from PDF and reformatted into sentence-per-line JSONL for NLP use; verse numbers, page headers, and footnote markers were programmatically removed. No wording was altered.
  • —This adaptation is an independent NLP dataset preparation and is not endorsed by the original translators or Seed Company.

Data Fields

Each line is a JSON object:

json
{"text": "<Tarok sentence>", "meta": {"source": "...", "license": "...", "attribution": "..."}}

Known Limitations

  • —Text is drawn entirely from the New Testament, so vocabulary and sentence structure skew toward religious/narrative register rather than everyday conversational Tarok.
  • —One entry (out of 3,699) contains a residual PDF-extraction artifact (a stray character) not yet manually corrected.
  • —Verse-level splitting means some entries are single clauses while others span several sentences, depending on the original verse length.

Contributions

If you would like to contribute additional Tarok text, audio, or corrections, please open an issue or discussion on this repository, or reach out via SeekSharp Labs.