Closed hank19960918 closed 3 years ago
This error happens when the alpha arg is equal to 1.0 of cutblur. Though this is definitely the implementation bug, we found that alpha as 0.7 shows the best performance so haven't fixed this.
Thanks for your immediate reply. I can run the code after changing the alpha value.
Thanks for your sharing, I met a problem when I ran the training code. My comment is like the following python main.py --model EDSR --augs cutblur --dataset DIV2K_SR --div2k_range 1-800/801-810 --scale 2 --dataset_root ./dataset/DIV2K. Thanks again