datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
osm-polygon-selection
osm-polygon-selection dataset
A curated set of OpenStreetMap polygons from 310
geographic units — sovereign countries plus sub-country regions
like Brazilian states, Chinese provinces, Indian zones, US states,
Canadian provinces, Japanese regions, and Indonesian islands —
classified by size bin (small / medium / large, area in
[0.1, 100] km²) and tagged by continent (Natural Earth admin0 lookup).
Size bins:
small — area in [0.1, 1) km² (10,000 m² to 1 km², roughly
100 m × 100 m… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-selection.osm-polygon-wikidata-only
OSM Polygon Wikidata, Wikipedia and Wikivoyage
OSM polygons carrying wikidata=*, enriched with multilingual Wikipedia and Wikivoyage documents. The published tables preserve regional records and provenance.
Source code: GitHub repository.
Dataset snapshot
Metric
Value
Polygon rows across regional extracts
1,184,110
Unique polygon identities (osm_type, osm_id)
1,157,841
Polygons with successful non-empty text (unique OSM identities)
650,663… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-wikidata-only.osm-polygon-website-tag
OSM Polygon Website Dataset
OpenStreetMap closed ways and polygon relations carrying a non-empty website OR contact:website tag, with full main-page text extracted using Trafilatura. Every statistic below is regenerated from the current upload-acknowledged Parquet artifacts.
Snapshot
Metric
Value
What it means
Snapshot status
In progress
Current published snapshot
Regional PBFs included
386 / 386
Published source shards / expected source PBFs… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-website-tag.osm-polygon-description-tag
OSM Polygon Description Tag
OpenStreetMap polygons with a successfully extracted trimmed non-empty
description or
description:<suffix> tag, published as one GeoParquet file per regional PBF
extract. Every row retains the complete original tag map, full Polygon or
MultiPolygon geometry, WGS84 geodesic area, bounding box, and OSM provenance.
Source repository: github.com/NoeFlandre/osm-polygon-description-tag.
Explore the pipeline metrics in the Trackio dashboard.
Read the… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-description-tag.fineweb-polygons
FineWeb Polygons
GitHub repository · Hugging Face dataset
The repository currently preserves one complete research line as Direction 1:
FineWeb polygon retrieval.
It contains the V1–V10 experiment history and is frozen as a baseline for a
separate future direction.
The public dataset catalog maps every immutable V1–V10 file to its country split, standalone version README, and reproducibility manifest. The readable GitHub catalog and HF metadata README link back to both projects.… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/fineweb-polygons.osm-polygon-wikidata-sentence-relevance
OSM Polygon – Wikidata Sentence Relevance
This dataset contains sentence-level, model-generated relevance annotations for
OSM polygons joined to Wikipedia and Wikivoyage sections. The labels are not
ground truth; validate them before use.
Public release map
Release
Contents
Card
Dashboard
V1 Afghanistan
54,462 two-question annotations, 161 polygons, 115 languages
v1-afghanistan/ · card
Afghanistan Trackio
V2 worldwide
200,000 H3-stratified binary… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-wikidata-sentence-relevance.put_polygonThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": null,
"total_episodes": 50,
"total_frames": 18444,
"total_tasks": 1,
"total_videos": 300,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 33,
"splits": {
"train": "0:50"},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/Jinyu220/put_polygon.osm-polygon-image-tag
OSM Polygon Image Tag
This dataset connects OpenStreetMap polygons to image references found in their
tags. It contains closed ways and relations only.
The release has three files:
polygons: one current row per OSM type and ID, including its geometry and
all original tags.
images: one row for each unique provider image. It keeps the provider's
reference and, when available, a direct image URL.
polygon_images: the links between features and images. This is a separate
file. One… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-image-tag.osm-polygon-web-search
OSM Polygon Web Search
This public Hugging Face dataset contains one explicitly approved POC table at
train.parquet, with one row per non-empty SAT sentence classified as
relevant-only by the local LiquidAI/LFM2.5-2.6B model. It is designed for the
Hugging Face Dataset Viewer: the table exposes polygon geometry, the exact
Brave query, page URL, title, HTTP status, full page text parsed by Trafilatura,
the SAT sentence, and the strict relevance result. The complete yes/no… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-web-search.osm-polygon-description-tag-worldcover
osm-polygon-description-tag-worldcover
A supervised text to land-cover dataset. Each example pairs OpenStreetMap description and localized-description tag text
with the ESA WorldCover class that covers at
least 80% of the OpenStreetMap polygon the text describes.
76,601 examples, 74,759 distinct polygons,
76,601 distinct documents.
from datasets import load_dataset
ds = load_dataset("NoeFlandre/osm-polygon-description-tag-worldcover")
print(ds["train"][0]["text"][:200]… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-description-tag-worldcover.osm-polygon-description-tag-eunis
OSM Polygon Description Tag
OpenStreetMap polygons with a successfully extracted trimmed non-empty
description or
description:<suffix> tag, published as one GeoParquet file per regional PBF
extract. Every row retains the complete original tag map, full Polygon or
MultiPolygon geometry, WGS84 geodesic area, bounding box, and OSM provenance.
Source repository: github.com/NoeFlandre/osm-polygon-description-tag.
Explore the pipeline metrics in the Trackio dashboard.
Read the… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-description-tag-eunis.osm-polygon-website-tag-eunis
OSM Polygon Website Dataset
OpenStreetMap closed ways and polygon relations carrying a non-empty website OR contact:website tag, with full main-page text extracted using Trafilatura. Every statistic below is regenerated from the current upload-acknowledged Parquet artifacts.
Snapshot
Metric
Value
What it means
Snapshot status
In progress
Current published snapshot
Regional PBFs included
386 / 386
Published source shards / expected source PBFs… See the full description on the dataset page: https://huggingface.co/datasets/NoeFlandre/osm-polygon-website-tag-eunis.put_polygon2This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": null,
"total_episodes": 50,
"total_frames": 19775,
"total_tasks": 1,
"total_videos": 300,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 33,
"splits": {
"train": "0:50"},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/Jinyu220/put_polygon2.