Interpolated data and test submissions

I saw that an interpolated version of the data (with equal voxel spacing) has now been provided as a branch on the github repository.
If I use that data to train and test my algorithm, would I able to submit these predictions to the challenge server for evaluation? Or would I have to resample them to original version of the data before submitting?

You may submit predictions in the spacing of either the master or interpolated branches. The evaluation system will recognize which version you’ve submitted based on the dimensions of the predictions that you submit.

Sounds good. Thanks for the quick response.