Comment by zellyn

Is the "books3" dataset mentioned in the Pile paper the one that authors are suing over? The one that includes a whole bunch of popular and copyrighted material?