Dan Bochman

Hi @markub3327 did you ever manage to implement DDIM for continuous time? I am also interested in this

@markub3327 It would be great if you can share the implementation because I also tried something similar but I am unable to make this work - Or what you posted...

I understand why you're seeing this problem. I'll fix this today.

I've pushed a dev version for now in Python 3 with fixed data paths Also, the CPU/GPU versions are merged and will automatically select the devices according to your CUDA...

Looking forward to this @SangbumChoi thanks for doing this @amyeroberts you're a great code reviewer