Aynkader/Balochi-dataset

Balochi Latin Syáhag Parallel Corpus & Grammar Dataset

This repository contains a high-quality, verified linguistic dataset designed to facilitate the integration of the Balochi Latin script (Balochi Latin Syáhag) into machine translation platforms, localized neural networks, and open-source computational models[cite: 2, 3].

📊 Dataset Contents

• Balóchiay Látini Syáhagay Rahband (PDF Book): A formal orthographic reference manual detailing standard syntax, diacritic rules, and morphological structures[cite: 2, 3].
• More than 33,000 Sentences Dataset (CSV): A parallel corpus featuring bilingual sentence pairs mapped side-by-side for machine learning model alignment[cite: 3].

🛠️ Orthographic Rules & Constraints

When training tokenizers or language models on this dataset, engineers must adhere to the core linguistic constraints of the Balochi Latin Syáhag system:

• Excluded Characters: The letters V, F, Q, and X are strictly absent from this orthography and are never utilized in authentic vocabulary[cite: 2].
• Diacritics & Vowels: Standardized vowel accents (á, é, ó) are implemented systematically to ensure phonological consistency across varied Balochi dialects.

🚀 Purpose & Computational Application

This data is structured to easily integrate into open-source benchmarking frameworks, localization pipelines, and platform translation training models (such as Google Translate requests). The clean CSV translation pairings allow for direct computational tokenization with minimal preprocessing.

⚖️ License & Contact

This dataset is provided openly to support digital linguistic inclusivity[cite: 3]. For inquiries, collaborative expansion, or further linguistic validation of the corpus, please open an issue in this repository or contact the maintainer[cite: 3].