Failed containers in the final test phase - Grace period

Dear participants,

There have been multiple containers which ran successful during the preliminary test phase but failed for the final test phase. In all cases that we checked, this was very likely because of programming errors that did not account for certain special cases in the predictions.

Still, to avoid excluding these approaches altogether and honor your efforts in participating in this challenge, we have decided to keep the final test phase open until Mon, 06.09.2021, 11:59 am (CEST). Please note that this extension is just to fix the error and no further adaptation of the algorithm is allowed. Please note that likely no additional account verification will be possible until this grace period.

We reserve the right to disqualify approaches: - which do not have a prior submission on preliminary test phase or for which the algorithm behaves differently on the preliminary testphase before/after the submission - for which no publicly accessible link with a time stamp before Sept. 3, 11:59pm (CEST) (arxiv, gitlab/github) that links to the description of the approaches is provided - for which we expect any malicious attempt in trying to exploit this grace period

Kind regards, Katharina

Dear Katharina, I submited for the final test phase but got an the same error "The algorithm failed on one or more cases". I used this docker container for preliminary test phase without any error. I also tried the algorithm and it showed the outputs. What should be the problem?Can you showe the error message? Thanks Wang

Dear Wang,

you hit the timeout (2 hours) for some images, while other images ran fine (although at around 20 minutes not really fast).

I don't know if this might have been influenced by a generally high load on the system yesterday evening, but it might also be an endless loop or a inefficient implementation.

Best regards Marc

Hi Dr. Maubreville, I get an error even I solved the problem with previous container. Could you look into my last container please? Regards,

Dear Marc, I resubmitted this afternoon, but I found the current status is very strange, why the resubmission failed in the afternoon, and yesterday’s submission was successful, can you help me check it, what is the situation now, or the system? Display error?! Thanks Wang

Hi Wang,

we again ran the samples for which your algorithm timed out to see whether this was due to high system load – it seems that this was the case and I am happy to report that your algorithm ran through without any issues.

So your submission is counted :)

Kind regards, Katharina

Dear Katharina, Thanks. But I remember that I also submitted docker in the afternoon and ran it again, so you actually re-ran the docker I submitted yesterday, did you not run the docker submitted this afternoon? Why did the docker run in the afternoon be wrong?

Kind regards, Wang

Hi Wang,

that is correct – your "newer" container was cancelled since there was already one with the same algorithm running.

Best, Katharina

Dear Katharina, Thank you very much for your help, thank you very much for your hard work Best, Wang