-
hi,
i encountered this error: "RuntimeError: mat1 and mat2 must have the same dtype, but got Half and Float" when trying to upscale using LDSR. I am using the latest google collab to run i.e. rena…
-
Hi, this is a great work!
May I ask which method performs better in text-to-3d generation, SDS-bridge or NFSD? It seems that the paper doesn't mention this comparision.
Looking forward to hearing fr…
-
Hi, thanks for releasing this amazing project. From my best understanding, current demo code does not use DDIM inversion during inference as the x_T variable is None in the following line? If so, coul…
-
There is certain values for the `ddim_steps` parameter, for which the model crashes with the stack trace appended at the end.
- 3 => crashes
- 9 => crashes
- 27 => crashes
- 81 => gets changed i…
-
The `numpy` backend has a limitation on literal accessor not shared by other backends.
Consider the following:
```python
import gt4py.cartesian.gtscript as gtscript
from gt4py.cartesian.gtscri…
-
Hi, thanks for sharing your excellent work!
I just walked through the code base and noticed that during sampling you used timestamp t from 0 to 999 (see [here](https://github.com/voletiv/mcvd-pytor…
-
Hello, I'm a beginner and would like to ask if this program just speeds up the reasoning process is that right, training won't be using it
-
Hi Phil,
It seems that the classifier guidance is missing for DDIM sampling. The `cond_fn` is not used in the `ddim_sample` function.
https://github.com/lucidrains/denoising-diffusion-pytorch/bl…
-
**when it is 512 pixel resolution it works but when 720 or 768 i get below error**
What does this error means?
It is stuck here for like 8 hours. I had enabled Loose Cross-frame attention and …
-
(my code)
samples = sampler.decode(z_enc, c, t_enc, unconditional_guidance_scale=opt.scale)
stack trace w/o pytorch forward and grad calls
File "/home/model-server/stable-diffusion/ldm/models/diff…