Knowledge distillation ; Pre-Trained Language Models AND ML/NLP blog
Common descendants