使用介绍
另一个用于掌握漫画艺术风格的 Lora,包含单独或与其他 lora 组合制作漫画艺术的工具。
V2 现在称为 manga-master,是 V1 的初始漫画助手 Lora 的改进版,虽然还不完美,但比之前的版本有了很大的改进。
调用时请使用以下规则:
ComfyUI 将优先使用 3 个关键词提示;manga, monochrome & grayscale,模型和剪辑的 lora 强度均为 1
Automatic1111 将优先使用 <lora:manga-master-V2:1> 调用输入,同时使用提示manga, monochrome & grayscale以获得完整效果
触发词: | manga monochrome grayscale |
作品参照
BREAK looking at viewer, standing, body view, masterpiece, absurdres, anime, manga, monochrome, grayscale, line art
Negative prompt: (embedding:verybadimagenegative_v1.3:1), (embedding:negative_hand-neg:1), (embedding:easynegative:0.5), (embedding:bad_prompt_version2-neg:1), text, text bobbles, comic
Steps: 30, Model: Anime\something_v22, width: 512, height: 768, Sampler: Euler a, CFG scale: 7, modelIds: , versionIds:
BREAK looking at viewer, standing, body view, masterpiece, absurdres, anime, manga, monochrome, grayscale, line art, laughing
Negative prompt: (embedding:verybadimagenegative_v1.3:1), (embedding:negative_hand-neg:1), (embedding:easynegative:0.5), (embedding:bad_prompt_version2-neg:1), text, text bobbles, comic
Steps: 30, Model: Anime\something_v22, width: 512, height: 768, Sampler: Euler a, CFG scale: 7, modelIds: , versionIds:
BREAK looking at viewer, standing, body view, masterpiece, absurdres, anime, manga, monochrome, grayscale, line art, running
Negative prompt: (embedding:verybadimagenegative_v1.3:1), (embedding:negative_hand-neg:1), (embedding:easynegative:0.5), (embedding:bad_prompt_version2-neg:1), text, text bobbles, comic
Steps: 30, Model: Anime\something_v22, width: 512, height: 768, Sampler: Euler a, CFG scale: 7, modelIds: , versionIds:
建议模型
本内容及资源均来自网络,版权归原作者所有,仅供个人学习研究,请勿商用和侵犯他人肖像,如需商用请联系原作者。若原作者内容有侵权之处请速联系我们,我们将会在24小时内删除。