Request for detail about pre-train DDPM

I'd like to ask you, if I need to extend it to multispectral bands, do I just need to change the in/out channel in the ddpm-train.json to the number of multispectral bands, and put the multispectral image in the dataset folder, and set the path? I have great admiration for your work, and I would like to use it as a comparison algorithm and quote your article.