Datasets:
metadata
license: mit
task_categories:
- summarization
- text2text-generation
- text-generation
- text-classification
language:
- en
pretty_name: OnionNewsScrape
This is a dataset of Onion news articles:
Note
- The headers and body of the news article is split by a ' #~# ' token
- Lines with just the token had no body or no header and can be skipped
- Feel free to use the script provided to scape the latest version, it takes about 30 mins on an i7-6850K