Dear participants,
Starting around 15:00 CET today, jobs stopped being queued for evaluation and remain stuck in the 'executing' phase, due to unexpectedly high traffic on Grand Challenge. We contacted the Grand Challenge team at 16:00 CET to see if they could assist us. However, they aren't able to resolve this problem as there isn't a straightforward way for them to prioritize jobs of a certain challenge.
We can't extend the grace period indefinitely due to the MIDL short paper deadline, which is tomorrow October 10th at 23:59 UTC-12. Those who submitted early deserve to know their results on the test set and their position on the leaderboard to finish writing their paper on time. However, it also doesn't seem fair to us if only three of the twelve hours of the grace period we decided on were usable due to technical problems on the platform.
We have decided on a compromise that will hopefully be acceptable to everyone. We will extend the grace period by another 12 hours, allowing submissions until tomorrow at noon, April 10th 12:00 CET.
However, for those who've already submitted, we'll be sending out the results of your models on the test set directly through the messaging system. This way, you can start incorporating these findings into your papers. We will also release the baseline model on the test set leaderboard so you can compare your model's performance against it. Be sure to check back tomorrow at 13:00 CET for your final leaderboard standings.
For Those Still Working on Submissions:
Please ensure your container is functioning correctly on the development leaderboard before submitting to the test set. Once you have checked this, submit to the test set as soon as possible. This way your job can get ahead in the queue of jobs that are still waiting to be executed. Jobs which haven't finished evaluating before 13:00 CET tomorrow will be excluded from the final ranking, so ensure you submit your test phase job ideally a couple of hours beforehand.
