metadata

dataset: fr-en-tupuri-translation
language:
  - fr
  - en
  - tpi
multilinguality: translation
license: mit
task_categories:
  - translation
task_ids:
  - machine-translation

French-English to Tupuri Translation Dataset

Dataset Summary

This dataset provides sentence-level translations from French and English into Tupuri, a low-resource language spoken in northern Cameroon and southwestern Chad. It aims to support research in African language translation and NLP for underrepresented languages.

Supported Tasks and Leaderboards

Machine Translation: French/English → Tupuri.
Multilingual modeling: Useful for zero-shot or low-resource model training.

Languages

fr: French
en: English
tpi: Tupuri (ISO 639-3: tur)

Dataset Structure

The dataset is in Excel format with the following columns:

french_sentence	english_sentence	tupuri_sentence
Merci	Thank you	Miindo

Data Splits

All data is currently in one file. Manual splits into train/val/test are recommended.

Usage

from datasets import load_dataset

dataset = load_dataset("your-username/fr-en-tupuri-translation")
print(dataset["train"][0])


@dataset{fr_en_tupuri_translation_2025,
  title     = {French-English to Tupuri Translation Dataset},
  author    = {Nopole Flairan, Ngana Noa},
  year      = {2025},
  publisher = {Hugging Face},
  note      = {Low-resource language dataset for Tupuri translation.}
}