Hi,
Thanks for sharing the codebase for the paper.
Despite your hard work on the project, it would be perfect if you could provide more details on how we can find the datasets and how they are preprocessed. Without this, it is a bit uncomfortable to reinvent the wheel.
I hope you could do this favour for the broader community, which is keen to use this method for further research.
Many thanks!