ml-cread
- This is the source code of the paper [CREAD: Combined Resolution of Ellipses and Anaphora in Dialogues](https://arxiv.org/abs/2105.09914). In this work, we propose a novel joint learning framework of modeling coreference resolution and query rewriting for complex, multi-turn dialogue understanding. The coreference resolution [MuDoCo](https://github.com/facebookresearch/mudoco) dataset augmented with our query rewrite annotation is released as well.