# Usage

            ## Load with `datasets`
            ```python
            from datasets import load_dataset

            ds = load_dataset("Boredoom17/roman-nepali-corpus")
            print(ds)
            ```

            ## Load directly with pandas
            ```python
            import pandas as pd

            df = pd.read_parquet("hf://datasets/Boredoom17/roman-nepali-corpus/nepali_corpus_roman.parquet")
            print(df.head())
            ```

            ## Suggested Workflow
            - Inspect `README.md` first.
            - Review `LICENSE.md` before redistribution.
            - Use the row-level metadata to filter by domain, script, or source.