CatDroid

What methods does ComfyUI have to iterate over each suggestion word (e.g., M words) and each reference image (e.g., N images) to obtain M*N generated images?

``` class CR_LoadImageList: ... images = torch.cat(image_list, dim=0) images_out = [images[i:i + 1, ...] for i in range(images.shape[0])] return (images_out, show_help, ) ``` why torch.cat and then un-pack to images_out...