Dorothy

This was a little experiment done for a friend to create a LoRA for one of his OC's. I'm posting it here because he does not currently have the ability to generate images offline. This way he can use Civit's image generator with the lora for his OC.

I'm posting 2 versions because he is used to using more booru inspired tags for prompting and I usually use e621 tags to prompt. So I trained the dataset on both styles. I don't want to make him have to switch over to my prompting style just to use this model but I'll post both so he can use them if he wants XD