Tweeted By @seb_ruder
Principle 2. How an algorithm scales is more important than its starting point. Avoid performance ceilings. Deep Learning is successful because it scales so effectively.
— Sebastian Ruder (@seb_ruder) September 13, 2018
Principles are meant to be controversial. I would argue that sample efficiency is at least as important. pic.twitter.com/jtjHmOcKlS