Open GZHU-DVL opened 1 year ago
still waiting for google to provie, stay tuned.
Hi, I update the code for simulation for now.
I wrote an unofficial implementation of UFO on the DiT imagenet, but the behavior is different from the text2image generation, which only supports 4 steps of generation.
Have a good job! Can you share the code of the paper "Semi-Implicit Denoising Diffusion Models (SIDDMs)"?