Powerful unsupervised domain adaptation method for dense retrieval. Requires only unlabeled corpus and yields massive improvement: "GPL: Generative Pseudo Labeling for Unsupervised Domain Adaptation of Dense Retrieval" https://arxiv.org/abs/2112.07577
I recon this is more of a generic question for TSADE + GPL (or any transformer used) , but can you use GPU by simply doing something like gpl.to(device)?
I recon this is more of a generic question for TSADE + GPL (or any transformer used) , but can you use GPU by simply doing something like gpl.to(device)?