The difference between Loss( f(x0+t_{n+1}*z) , f(x0+t_{n}*z) ) and Loss( f(x0+t_{n+1}*z), x0 )? about consistency_models HOT 4 OPEN

I think this might just make it a score-based model. Like Yang Songs earlier work, or instead very like EDM. The +1 wouldn't really matter I feel.

from consistency_models.

Or is it possible to use both Loss( f(x0+t_{n+1}*z), x0 ) and Loss( f(x0+t_{n+1}*z) , f(x0+t_{n}*z) )? The first term enforces the similarity with the ground truth images, while the second term enforces the consistency. Simply using the consistency term would in fact decrease the generation performance or lead to longer training.

from consistency_models.

loss(x0+t_{n+1}*z), x0) can only enforce the model to learn the expectation of x0, not the target point of the current trajectory. It is a wrong loss term for one step diffusion

from consistency_models.

loss(x0+t_{n+1}*z), x0) can only enforce the model to learn the expectation of x0, not the target point of the current trajectory. It is a wrong loss term for one step diffusion

Yes, although you can target the GT I think this would actually make the task harder. By trying to target the trajectory you are smoothing the data manifold which makes training easier.

from consistency_models.