1 paper
Longteng Zhang, Sen Wu, Shuai Hou +7
Adapting large pre-trained language models to downstream tasks often entails fine-tuning millions of parameters or deploying costly dense weight updates, which hinders their use in…