Test mode - why don't you evaluate the loss function ?

[LulaSan]

Hi! I was trying the test mode in your code "experiments.py" and I figured out that you don't obtain as output the loss function, in order to see if it is the same as the one in training time.
I have seen in the paper of monodepth that they only obtain disparity, convert into depth and compare with GT.
Why don't you calculate also the loss function?

Thank you,
Luisa