Fetching the paper…

Adapt-and-Distill: Developing Small, Fast and Effective Pretrained Language Models for Domains · Around