Datasetpublic

kalomaze/md-2-xml-wiki-tables

md-2-xml-wiki-tables 958 markdown tables extracted from fan/community MediaWiki sites for markdown-to-XML format conversion tasks. Format JSONL with fields: title: article title from the source wiki page section: section heading the table appeared under wiki: source wiki name table_md: raw markdown table filename: original filename Splits train: 894 tables eval: 64 held-out tables Source Various fan/community MediaWiki sites. Most… See the full description on the dataset page: https://huggingface.co/datasets/kalomaze/md-2-xml-wiki-tables.

sourceHugging Facecc-by-nc-sa-4.0updated 8mo agoView on Hugging Face

1likes28downloads

Dataset Card

md-2-xml-wiki-tables

958 markdown tables extracted from fan/community MediaWiki sites for markdown-to-XML format conversion tasks.

Format

JSONL with fields:

—title: article title from the source wiki page
—section: section heading the table appeared under
—wiki: source wiki name
—table_md: raw markdown table
—filename: original filename

Splits

—train: 894 tables
—eval: 64 held-out tables

Source

Various fan/community MediaWiki sites. Most use CC-BY-SA, some use GFDL or CC-BY-NC-SA.

Usage

Used with the `kalomaze/md-2-xml` verifiers environment for training and evaluating models on structured format conversion (markdown table to XML).