To clarify, I retouch them with photoshop (fix extra fingers and things like that), but photoshoped images do not have metadata so I add the metadata from the original unedited generation.
So well, I just wrote the original commentary. Not pretending to create some literary masterpiece. Originally, I wanted to create an animal pov image with a cat stretching in front of her first person pov. Problem is AI is not on that level yet (and the animal_pov tag is so underused in imageboards it's nearly impossible to create a lora for that, lol.) That's my way of compensating what I can't physically do.
