This LoRA has been tested and can generate amazing avatars on Lucid's MIX and Crosskemono(F/E/2.0) models,It may not work properly on other models.
Tips:
>>>>>>>>>>>>>> Before using this Lora <<<<<<<<<<<<<<
>>> Please set "Settings - Stable Diffusion - Clip skip" to 2 <<<
>>>>>>>>>> if you want to generate an avatar, <<<<<<<<<<
>>>>>> please be sure to add "portrait" in the prompt.<<<<<<
>otherwise this LoRA will not be able to generate the best image.<
First mixed stable build release