Google’s open source NLP model uses 1.6T parameters to finish pre-training in a quarter of the time February 18, 2021 // by cascpix This content is for members only. Sign up for access to the latest trends and innovations in fintech. View subscription plans. Login