Efficient NLP policy document

The amount of computation put into training NLP models has grown tremendously in recent years. This trend raises the bar for participation in NLP research, excluding large parts of the community from experimenting with state-of-the-art models. It also creates environmental concerns since this computation uses increasing amounts of energy. Attached you can find a document which is the report of a working group appointed by the ACL Executive Committee to promote ways that the ACL community can reduce the computational costs of NLP and thereby mitigate some of these concerns. The recommendations in this report are guided in part by the results of a survey we conducted with the ACL community as well as a presentation of the main ideas during the ACL 2021 business meeting. Before submitting our report to the ACL Executive Committee, we would like to kindly ask the community for any feedback on the document using the following link https://forms.office.com/Pages/ResponsePage.aspx?id=9028kaqAQ0OMdrEjlJf7.... We will accept feedback until Friday, November 19th.

Best regards,
Yuki Arase, Phil Blunsom, Mona Diab, Jesse Dodge, Iryna Gurevych, Percy Liang, Colin Raffel, Andreas Rücklé, Roy Schwartz, Noah A. Smith, Emma Strubell and Yue Zhang