> Detailed instructions for the final submissions have now been added on the wiki:
> * [**Instructions for final submissions**](syn6156761/wiki/406537)
> * [**Write-up template**](syn6156761/wiki/406631)
In many challenges there is a drop in participation for the final phase. Please do not be discouraged from making final submissions if you were not among the top teams in the leaderboards:
* **Team rankings often change substantially on the final hold-out data**
* **You may still be included in the consortium and the paper even if you are not a top performer**
* **Only teams with final submissions will get pre-publication access to challenge data and results**
* **A principle of DREAM is "do no harm"**: in our analysis we focus on the top performing methods, we will not discuss individual methods that performed poorly. There can be many reasons for a low performance, including a simple mistake or sub-optimal strategy to tune the parameters.
Created by Daniel Marbach daniel.marbach Hi Suresh
Do you still need to make new submission or your submission for subchallenge2 is the final? if you need to change your submission please send an email to me.
Hello Daniel,
I have submitted by accident the template zip file of subchallenge1 instead of my results zip file. Is there a way to cancel this submission and I submit my results? The website says that I have reached my quota.
Thank you very much,
Jose I just figured that I had made an inadvertent submission to the subchallenge 2 final queue a week earlier. Is there a way to invalidate that and upload another zip file?
Thanks,
Suresh I made the final submission for subchallenge 1, but I am unable to make the final submission for subchallenge 2. The system reports that I have exceeded my quota of 1 submission. Could someone look at this?
Thanks,
Suresh You mean because of the last point in _Section 3. Code_? We just "recommend you keep track of random seeds", but it's not a strict rule and methods do not have to be deterministic.
--daniel hi, daniel, i saw you said non-deterministic methods are not allowed, but we were too lazy to implement anything, we just called a third party software. that software is not deterministic, and produces results that can be 30% difference even in the number of modules found every round. we are just trying on leaderboard to see if any of it happens to have a high score.