Datasets:

ccdv

/

arxiv-summarization

like 68

Greetings!

First off, awesome dataset for summarization training.

The dataset card says there is a field for the 'id' for the paper id. When I look at the features for the train split I see "{'article': Value(dtype='string', id=None), 'abstract': Value(dtype='string', id=None)}". Or, if I look at the keys directly I only see article and abstract. Is there a version with paper id?

Thanks!
-Andrew