Myanmar Raw Sentences Corpus (datalamhnyin/myanmar-raw)

A collection of Burmese text sentences compiled from various public sources, intended for Myanmar NLP and language model pre-training research.

Status (Raw/Uncleaned):
This dataset is in its raw stage. Basic deduplication and length filtering have been applied, but it may contain OCR noise, non-standard spelling variations, or encoding artifacts. Further task-specific normalization is recommended before deployment.

---

Dataset Overview

• Total Rows (Sentences): 1,625,401
• Language: Burmese (Myanmar)
• Format: Parquet (Single column: text)

---

Acknowledgements

Special thanks to Khant Sint Heinn (Kalix Louis), Machine Learning Engineer specializing in NLP, for his technical advice, constructive feedback, and valuable discussions regarding pipeline workflows and text processing strategies.

---

Disclaimer & Intellectual Property Notice

• Ownership & Operation: This repository is independently created, maintained, and operated solely by DataLamHnyin.
• Liability: All activities related to data collection, processing, and hosting are the sole responsibility of DataLamHnyin. Advisors or contributors acknowledged above bear no legal liability, affiliation, or ownership regarding the underlying contents or sources of this dataset.
• Takedown Requests: This dataset is provided strictly for academic and non-commercial research purposes. If you are a copyright holder and believe that any material should be removed or attributed differently, please submit a request via the Hugging Face Community tab or contact DataLamHnyin directly.