Evaluation code and the dataset

Hi, thank you for the great work and for open-sourcing the code!

I was wondering if there are any plans to release the evaluation code and the dataset (or at least the test split annotations) used in the paper. 

Thanks again!