LoRA Stable Diffusion Low Rank Adaptation Models itsuki-lora:Both v1 and v2 trained with YesMix v1.5, v1 is overcooked and sucks, don't even use it, v2 works fine.