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!
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
andlatest_net_s_gen.pth
, how should I modifymodel_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!