Datasetpublic

OpenMLRL/arXiv_abstract

arXiv Abstract This dataset is based on the arXiv scientific papers and is used for the text expansion task. (Download raw data here). I processed the raw data for the article expansion task with extract_arXiv_abstract.py. The processed dataset only contains the article ID and abstract fields, and the abstract length should be 100-300 tokens. The JSON objects are in the following format: { 'article_id': str, 'abstract_text': List[str], 'token_count': int }

sourceHugging Faceupdated 1y agoView on Hugging Face

0likes39downloads

Dataset Card

arXiv Abstract

This dataset is based on the arXiv scientific papers and is used for the text expansion task. (Download raw data here).

I processed the raw data for the article expansion task with extract_arXiv_abstract.py. The processed dataset only contains the article ID and abstract fields, and the abstract length should be 100-300 tokens. The JSON objects are in the following format:

{ 
  'article_id': str,
  'abstract_text': List[str],
  'token_count': int
}