# Roman Nepali Corpus Overview

            Rows: 307,999

            This subset isolates Latin-script Nepali from the colloquial corpus.

            ## Repository Contents
            - `README.md`: dataset card
            - `LICENSE.md`: license summary
            - `docs/OVERVIEW.md`: high-level summary
            - `docs/DATA_SCHEMA.md`: field descriptions
            - `docs/DATA_SOURCES.md`: provenance notes
            - `docs/USAGE.md`: loading examples
            - `docs/QUALITY_AND_LIMITATIONS.md`: dataset caveats
            - `CITATION.cff`: citation metadata
            - `nepali_corpus_roman.parquet`: primary parquet file