Fetching the paper…

Mixout: Effective Regularization to Finetune Large-scale Pretrained Language Models · Around