In both challenges, the data is split into train/test/eval sets.
In the first stage of the the competition, the train/test sets are provided
for training.
They contain only the ground-truth shapes, Y.
The partial data, X, is generated by the participants.
Some example routines are provided.
Custom routines may be used if reported.
At a later stage, the eval set is provided to generate the submission for the
final evaluation.
In this set, the partial data, X, is generated by the organisers and provided
to the participants.
The corresponding ground-truth shapes, Y, are kept secret.