Dear participants,
We hope you are doing great. This is a friendly reminder that the deadline for uploading your submissions to the DA-VINCIS shared task is fastly approaching. Also, we would like to share with you all that a couple of days ago a participant informed us of potentially mislabeled samples in the training set. We already verified this, and found that a small subset of samples may be mislabeled in the training set (less than 3%). We apologize for this inconvenience and will perform a post-challenge analysis on the impact of such noisy examples on the performance of baseline models.
Looking forward to receiving your submissions!
Best
Organizing team
Posted by: hugojair @ May 17, 2022, 4:06 p.m.