About training

[windfollowingheart]

Could you proivde the detailed train parameters, the proportion of train and test dataset, and the train order of the defocusnet, raftnet and fusionnet. I have try to train , but dont konw how to get the good results. Thank you.

[qlfwake]

Did you find the "Experiment" file, I tried to run "train" but did not see this file in "from sacred import Experiment"