OpenMatch/cocodr-large

This model has been pretrained on BEIR corpus without relevance-level supervision following the approach described in the paper COCO-DR: Combating Distribution Shifts in Zero-Shot Dense Retrieval with Contrastive and Distributionally Robust Learning. The associated GitHub repository is available here https://github.com/OpenMatch/COCO-DR.

This model is trained with BERT-large as the backbone with 335M hyperparameters.