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
Previous code does not detect OOM error in QGen, which might be due to large QPP or batch size
modified: gpl/toolkit/qgen.py
: Added try catchnew file: tests/unit/test_qgen.py
: Added test