Dear authors,
We are comparing with your method and wanted to confirm if you train on the official train split of RE10K to ensure fair comparison.
Additionally, where can we find the list of training datasets that you used for the model reported in the paper.
Many thanks!