wjosielct/shawi-spanish-corpus

The Shawi-Amazon Corpus 🌳

A parallel dataset for Shawi (Chayahuita) - Spanish machine translation.

• Source: Yosë nanamën (New Testament + Genesis) © Wycliffe Bible Translators.
• Target: Spanish (Reina Valera 1909 & Biblia Libre para el Mundo).
• Domain: Religious / Biblical.
• Size: ~9,200 sentence pairs.

Dataset Structure

1. Master File (shawi_spanish.tsv)

Contains metadata (Book, Chapter, Verse, Split) and columns for both Spanish variants.

2. Moses Format (splits/)

Plain text files ready for training (Fairseq, Moses, OpenNMT).

• shawi_spanishRV1909: Shawi to Standard Spanish.
• shawi_spanishBLM: Shawi to Modern Spanish.

Includes train, dev, and test splits.

Usage

from datasets import load_dataset

# Load the dataset (requires defining a loading script or using the files directly)
# For now, you can download the files manually or use:
dataset = load_dataset("wjosielct/shawi-spanish-corpus", data_files="shawi_spanish.tsv")

Citation

If you use this dataset in your research, please cite the following paper and dataset:

Paper (ResearchGate): https://doi.org/10.13140/RG.2.2.27497.56169

Dataset (Zenodo): https://doi.org/10.5281/zenodo.18259277

@dataset{corbera_terrones_2026_18259277,
  author       = {Corbera Terrones, Josiel},
  title        = {Shawi-Amazon: A Parallel Dataset for Spanish-Shawi
                   Low-Resource Machine Translation
                  },
  month        = jan,
  year         = 2026,
  publisher    = {Zenodo},
  version      = {1.0.0},
  doi          = {10.5281/zenodo.18259277},
  url          = {https://doi.org/10.5281/zenodo.18259277},
}