Hi,

thank you for releasing the code and dataset.
Will you also release the pretrained weights of your model that you have used for your leaderboard submission?
This would make it easier for people to reproduce the results/evaluate TCP on other benchmarks.