Closed foolhard closed 1 year ago
Unfortunately, that is not possible as we need to limit the number of times participants are able to evaluate their models both on the validation and test set.
hi, @foolhard , have you been able to reproduce the baseline results for task2?
@Lavreniuk I guess all problems were solved with the new post-processing method. I will thus close the this issue.
Yes, I can submit the results to the leaderboard correctly with the new post-processing code. Thanks for your kindly support.
Hi,
Could you release the metric code for Tasks 1 and 2? It's convenient for us to test the val set locally.
Thanks.