Throws error if no source image is added. But why? Couldn't you just use a custom lora model, and provide a prompt? Does it have to be img2img?
· Sign up or log in to comment
