will it be possible to finetune the sd2.0 depth model?

[xdfeng07]

If sd2.0 depth model could be used to generate images of designated learned characters， while using input image as guidance for pose and scene, that will really help a lot

[markusmalmros]

Yes, this would be very useful.