This dataset is configured so the Hugging Face Dataset Viewer loads qna.csv as the primary data file.
The repository also contains all_links.csv, which has a different schema and is kept as an auxiliary file rather than part of the default viewer configuration.
This avoids schema-casting errors caused by the Hub trying to combine both CSV files into one split.
