A PyTorch Implementation of Image Super-Resolution via Iterative Refinement (https://arxiv.org/abs/2104.07636)
An explanation of this can be found at https://medium.com/@adityanutakki6250/sr3-explained-and-implemented-in-pytorch-from-scratch-b43b9742c232
You can train on your custom data, by pointing the 'dataset' module to wherever your data is located. change the target resolution to whatever you like