stable-diffusion icon indicating copy to clipboard operation
stable-diffusion copied to clipboard
verified publisher icon CompVis

Confused with nll_loss

Open haichao2023 opened this issue 1 year ago • 2 comments

Is this correct? Sum all pixels and divide by batch_size? https://github.com/CompVis/stable-diffusion/blob/21f890f9da3cfbeaba8e2ac3c425ee9e998d5229/ldm/modules/losses/contperceptual.py#L57

haichao2023 avatar Nov 26 '24 04:11 haichao2023

+1, I am also confused about this loss

onlyonewater avatar Dec 09 '24 08:12 onlyonewater

also confused, anyone already train a VAE result ? is it OK ?

ljw919 avatar Jan 13 '25 12:01 ljw919