Updates on test set evaluations

In line with the extended submission deadline, we also extend the number of possible test set evaluations to 20 for each task. This upper limit will not further be extended.
In addition, we discovered a misconfiguration in eval.ai that allows submissions to be flagged as private. We apologise for this.
Evaluation results (also if they were initially marked as private) will be published on the leaderboard in order to make comparison between teams possible. Going forward, please make sure to use the “public” option for your evaluations.