Will they available on the private dataset?
Hi @golovanov, the models should not use any labels as the competition is about the unsupervised representation learning. The models will be trained on randomly permutated datasets on the evaluation servers.
1 Like