I loved this paper: https://arxiv.org/pdf/2106.07465.pdf and I'd love to add the dataset they are using to our benchmark.

However, I'm unsure how to use the physics library required to generate the data and would rather just have a simple CSV with the data (csv columns can contain arrays if need be, but I don't think this will be required here).

Feel free to PR this csv into https://github.com/mindsdb/benchmarks

This will count 3 points towards the hacktoberfest deep learning laptop raffle.