Compact Speech Translation Models via Discrete Speech Units Pretraining

Open in new window