docs/USAGE.md

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.