cnfs-super-resolution icon indicating copy to clipboard operation
cnfs-super-resolution copied to clipboard

Published 20 hours ago • verified publisher icon christina-winkler

forward error!!!

Open qinglew opened this issue 3 years ago • 1 comments

In line 62 of optimization/trainer_baseline.py, the return values of model.forward(y, x) is z, x_bpd represents the latent code and minus log-det respectively according to class FlowModel in models/architectures/condNF.py. However, you receives the return values with logp_mass, _, _. It's an error obviously.

qinglew avatar Nov 22 '21 11:11 qinglew

Hi,

Thanks for pointing out this error. I will try to fix it in due time. Feel free to do a pull-request!

christina-winkler avatar Nov 22 '21 17:11 christina-winkler