Source-Free Domain Adaptation for Question Answering with Masked Self-training

1 code implementation • 19 Dec 2022 • M. Yin, B. Wang, Y. Dong, C. Ling

In this study, we investigate a more challenging setting, source-free UDA, in which we have only the pretrained source model and target domain data, without access to source domain data.

Question Answering Source-Free Domain Adaptation +1