US names dataset v0.2.0

US-wide synthetic-name dataset, built from pinned Census 2020 and SSA national data (1880–2025).

• Normalized, indexed SQLite tables for name frequencies, demographic counts, sex counts, and birth-year counts.
• Python generator supports consistent given-name sex selection, initials, multiple given names and surnames, length constraints, and weighted or uniform sampling.
• Download names.sqlite.gz and decompress it for use with the Python source at this release tag. tokens.jsonl.gz is the portable export.
• names.manifest.json records original URLs and SHA-256 hashes; SHA256SUMS verifies the release assets.

These are approximate synthetic distributions. Census demographic groups do not establish linguistic or cultural origin. Joint name, demographic, sex, and cohort relationships are approximated; source suppression and missing names limit coverage. Source-recorded sex is not an individual's gender identity. See README.md for limitations and defaults.