drboog / Shifted_Diffusion

Code for Shifted Diffusion for Text-to-image Generation (CVPR 2023)

drboog/Shifted_Diffusion Issues

IndexError: The shape of the mask [7] at index 0 does not match the shape of the indexed tensor [8, 1536] at index 0
Updated 5 months ago
RuntimeError: The NVIDIA driver on your system is too old
Updated 8 months ago
How to create multiple image embeddings from a single text embedding
Updated a year ago
two questions
Updated a year ago2
Great work! where can i get these files?
Updated a year ago
pytorch version
Updated a year ago1
How to get the mean and std of Gaussians of my dataset?
Updated a year ago
Generated images are not as sharp as SD
Updated a year ago
Question about COCO zero-shot evaluation.
Closed a year ago1
The cosine similarity of generated and gt image embeddings is not increasing during training
Closed a year ago8
how much memory do we need for training?
Closed a year ago8
About ckpt
Closed a year ago8
ckpt name seems not match
Closed a year ago6
Is there codes for the diffusion decoder as well?
Closed a year ago1