Datasets:

davidr99
/

us-address-standardization

license: mit
language:
  - en
task_categories:
  - text-generation
tags:
  - address
  - address-standardization
  - postgis
  - stdaddr
  - us-addresses
size_categories:
  - 10K<n<100K
configs:
  - config_name: default
    data_files:
      - split: train
        path: train.jsonl
      - split: validation
        path: validation.jsonl

US Address Standardization (PostGIS stdaddr schema)

Chat-format instruction data that maps a raw US address string to a strict JSON object matching the PostGIS address_standardizer stdaddr type (camelCase keys, USPS-abbreviated values).

Each row has three flat fields (system / user / assistant) for easy reading and grepping; rebuild the chat messages list from them at train time:

{
  "system": "<standardization instructions>",
  "user": "100 Old Forge Rd, Kent, CT 06757",
  "assistant": "{\"building\":\"\",\"houseNum\":\"100\", ...}"
}

Output keys (16, in order)

building, houseNum, predir, qual, pretype, name, suftype, sufdir, ruralRoute, extra, city, state, country, zipcode, box, unit

sufdir is the post-directional (e.g. NE in "Main St NE"); extra carries floor info or an intersection's cross street; ruralRoute is number-only.

Generation

Synthetic, label-first: a valid structured record is composed, then a deliberately messy raw string is rendered from it (varied casing, abbreviation spellings, punctuation). Covers streets, numbered routes, interstates, intersections, rural routes, PO boxes, weird building names, and tricky unit designators. See the scripts/ folder in the companion model repo for the generator. For exact PostGIS fidelity, labels can instead be produced with postgis_label.py against a live PostGIS instance.