-
Getting Error in the last cell during the launch:
WARNING[XFORMERS]: xFormers can't load C++/CUDA extensions. xFormers was built for:
PyTorch 2.2.1+cu121 with CUDA 1201 (you have 2.3.0+cu121)
…
mcqx4 updated
3 months ago
-
I have only just started exploring all this wonderful MLX stuff but when I try to run the `image2image.py` script on a simple `.png` I am getting an error that seems to arise during the `unet.py` resh…
-
I get the exact same results when using the draw mask on image-to-image. Is this a bug, or is it not supported? And if it's not supported, maybe the draw feature should be disabled for these?
I hop…
-
### Question
Is this process processing one image at a time ? Does ADetailer support running in batch mode? This will definitely improve the running speed, but I see that the code seems to only be a…
xddun updated
6 months ago
-
This is a feature question, probably reflecting my limited experience on the subject rather than an MLX limitation. I am trying to modify the MLX SD example to accept an input image, encode the image…
-
I try to use the dpm to accelerate guided diffusion sample, but the results is bad. the model link is https://github.com/Janspiry/Palette-Image-to-Image-Diffusion-Models/tree/main/models/
![000168](h…
-
Hi,
I'm training Dreambooth with dog dataset test, (Prior-reservation loss case) using HPC machine having Tesla V100 card, pytorch 1.10.1,diffusers 0.22.0.dev0, accelerate 0.20.3. The following is…
unrue updated
10 months ago
-
Hi Fangda,
Thanks for the great work!
I have some questions about the unsupervised image-to-image translation baselines, namely CycleGAN and CUT, reported in the paper (Table 1). I was wondering…
-
Function: increases resolution and sharpens images, also produces higher quality faces.
This thread appears to have the code necessary to reproduce this feature using diffusers:
https://github.com…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues and checked the recent builds/commits
### What happened?
Images below 0.7 Denoinsing strength generate blurr…