datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
indian-addresses-raw
Indian Addresses (Raw)
Raw, unstructured Indian address strings — the unlabeled corpus behind
gagan1985/qwen3-0.6b-indian-address-parser.
For the labeled/parsed version, see
gagan1985/indian-addresses-gold.
Sources
Two raw source types, distinguished by source_type:
mca — Indian Ministry of Corporate Affairs company-registration addresses
(public disclosure data; CIN numbers are part of India's public company registry)
bank — bank/business-correspondent branch… See the full description on the dataset page: https://huggingface.co/datasets/gagan1985/indian-addresses-raw.DoD-Instruction-8410-DNS-IP-Address-Use-And-Approval
🌐 DoD Internet Domain Name and IP Address Resource Question-Answer Dataset
Source: DoD Instruction 8410.01
Source Effective Date: December 4, 2015
Change Incorporated: Change 1, effective June 4, 2021
Source Organization: Office of the DoD Chief Information Officer
Source Ownership: United States Department of Defense
📋 Overview
Dataset Summary
The DoD Internet Domain Name and IP Address Resource Question-Answer Dataset is a structured… See the full description on the dataset page: https://huggingface.co/datasets/leeroy-jankins/DoD-Instruction-8410-DNS-IP-Address-Use-And-Approval.us-address-standardization
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). Used to fine-tune the qwen35-address-std model family.
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… See the full description on the dataset page: https://huggingface.co/datasets/davidr99/us-address-standardization.indian-addresses-gold
Indian Addresses (Gold Standard)
Span-labeled Indian addresses — the training data behind
gagan1985/qwen3-0.6b-indian-address-parser.
For the raw, unlabeled corpus, see
gagan1985/indian-addresses-raw.
4,834 records, split by provenance:
reviewer
Count
Description
llm:deepseek/deepseek-v4-pro
4,825
LLM-reviewed via OpenRouter, span-verified against the raw address text (a field value not found verbatim in the source string is dropped, not guessed)
gagan
9… See the full description on the dataset page: https://huggingface.co/datasets/gagan1985/indian-addresses-gold.us-address-comparison
US Address Comparison (per-part match verdicts)
Chat-format instruction data for judging whether two US addresses refer to the
same place, part by part. It is the companion to
davidr99/us-address-standardization:
each address is provided as the raw string plus its standardized JSON (the
output of the normalization model).
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":… See the full description on the dataset page: https://huggingface.co/datasets/davidr99/us-address-comparison.