模型简介: 不用controlnet,直接可以生成。 颜色倾向于红色,应该可以通过加其他颜色的权重来解决。 云跑,在一个小时内完成三个lora的测试着实有些困难。最后几张潦草的图就是证明。 ~~~~~~~~~~~~~~~~~~~~~~~o(* ̄︶ ̄*)o~~~~~~~~~~~~~~~~~~~~~~~ 测试提示词和参数: Prompt: (masterpiece:1.3),(best quality:1.2),8k,absurdres,(extremely detailed:1.3),highestres, (1girl:1.2),<lora:lowRA:0.6>, dancing, solo, dress, jumping Negative prompt: easynegative,bad-hands-5,lr,(monochrome:1.2),facial mark,facial tattoo,border,frame,(pantyhose),(low quality:1.3),bad artist's work,(crossed eyes:1.3),(mole:1.3), (NegfeetV2:1.2),sitting, Steps: 20, Size: 512x768, Seed: 2107131683, Model: henmixReal_v40, Sampler: DPM++ 2M Karras, CFG scale: 7, Clip skip: 2, Model hash: f4151d2b7b, Hires upscale: 2, AddNet Enabled: True, AddNet Model 1: dancer-10(aa05a5bfcd5a), Hires upscaler: R-ESRGAN 4x+, AddNet Module 1: LoRA, AddNet Weight A 1: 1, AddNet Weight B 1: 1, Denoising strength: 0.4 |
1人表示很赞