Hi, thanks for organising a great challenge and a brilliant dataset. I'm interested in benchmarking algorithms on the test set and comparing to the published leaderboard results. However, it seems that the post-challenge data release contains only fully-sampled test data (and val data), without the masks (or undersampled data) included. Therefore, it would be impossible to do a fair comparison to the published leaderboad, since the mask per patient varies (i.e. acceleration factor and sampling pattern), and random mask generation is not reproducible without a seed. Are the test set masks available? Or are the acceleration factors, sampling pattern names and seeds available to generate them? Thanks in advance!