Large corpus of uncompressed and compressed sentences from news articles.
google-research-datasets/sent_comp · main · files are served by the source, never re-hosted here
