Some questions regarding training visualizations and training models

[jiunyen-ching]

Hi!

I would like to ask:
Q1. For the training visualization, are the images arranged (from top to bottom) in the order of ...

color groundtruth image + mask
color structure image
color groundtruth image
mask
reconstructed image
... ?

Q2. Regarding training Structure and Flow models, after I finished training both Models 1 and 2 thus obtaining latest_net_f_gen.pth and latest_net_s_gen.pth, how should I modify model_config.yaml to resume training from those 2 weights?

Q3. Assuming that I have made the necessary changes to model_config.yaml in Q2, can I re-use the same training & val datasets (used to train Model 1 and 2) to do training for Model 3?

I hope you can help me to clear my doubts and questions. Thank you!