andreas

Setting `num_threads=0` solves it for me. Does that work for you as well? https://github.com/zdlarr/Location-aware-SIRR/pull/4/commits/ad52c07811151c7501ea5a1f132eea96d3daef5a

We describe how to modify the number of steps in the section "How to speed up the inference?" in the README.md. Does that help? What is the error you get...

I had the same problem. It seems to happen because of the upper cases. after copying and renaming the cmake files like the following I could compile. ``` ls cmake/Modules/...

Do I have to download a binary named "rectangle-props"? I could only use setup.py when I commented out this file. ``` 1 from setuptools import setup 2 from distutils.sysconfig import...

We use the same version as in the pip package. ```bash pip install lpips ``` ``` python import lpips loss_fn_alex = lpips.LPIPS(net='alex') d = loss_fn_alex(img_super_resolution, img_ground_truth) # normalized to [-1,...

We just released the code. It also contains the functions for the measurements. Hope this helps!

Thanks for your interest! Please try to use the following values in the [config file](https://github.com/andreas128/SRFlow/blob/205e63f1f96f8d98956c4a1034c6cfa04c6db8bc/code/confs/SRFlow_DF2K_4X.yml#L57). ``` in_nc: 1 out_nc: 1 ``` Does that solve you problem?

# How to change the number of output channels for SRFlow? To change the output dimension, you need to change the architecture of the Normalizing Flow. At the moment, the...

Thanks, so did the architecture change work for you?

Okay great! Then you will have 1x(2x2)=4 channels in the first scale level. This means that the conditional affine image injector will extract features from only 2 channels and applies...