rancper/wikipedia

Dataset Card for Wikimedia Wikipedia Dataset Summary Wikipedia dataset containing cleaned articles of all languages. The dataset is built from the Wikipedia dumps (https://dumps.wikimedia.org/) with one subset per language, each containing a single train split. Each example contains the content of one full Wikipedia article with cleaning to strip markdown and unwanted sections (references, etc.). All language subsets have already been processed for recent dump… See the full description on the dataset page: https://huggingface.co/datasets/rancper/wikipedia.