facebookresearch/ViewDiff
ViewDiff generates high-quality, multi-view consistent images of a real-world 3D object in authentic surroundings. (CVPR2024).
PythonNOASSERTION
Issues
- 4
The version issue of the transformers package.
#25 opened by zy2024bit - 0
ncclInvalidUsage error
#24 opened by xiyichen - 0
intergate viewdiff to huggingface hub
#23 opened by jadechoghari - 1
Training Time and Dataset
#21 opened by Yeez-lee - 9
train without projection layer
#17 opened by immaturehearts - 2
pipeline question / paper question
#16 opened by sararoma95 - 4
a question for training (with small GPU)
#15 opened by minounou - 1
Modifying a loaded unet model
#14 opened by junsukha - 3
Autoregressive generation when n_c = 1
#13 opened by junsukha - 3
- 1
Can't open the project website
#11 opened by yiboz2001 - 6
Nan loss during training
#10 opened by ItsThanhTung - 14
- 2
dataset split
#9 opened by yijicheng - 2
dataset split
#8 opened by ShengyuH - 4
- 1
Trained model
#5 opened by junsukha