This lora uses a large model "universal deliberate_v2" for training. It is suggested that when using this lora, the intensity control should be between 0.4-0.5;Sampling method: Euler a; Iteration steps: 35;In addition, it is recommended to use controlnet (openpose, canney, depth) for character control;I should think so! This lora can not only produce character plaster style, any other picture with controlnet can come out plaster style!I hope you give birth to your favorite plaster model! Have fun!!
