以下是关于 chilloutmix 模型的相关信息:
模型:chilloutmix_NiPrunedFp32Fix生成的尺寸和你的草稿图保持一致(cn垫图512x768,高清修复2倍,重绘幅度0.35)关键词和ControlNet参数每张不一样都进行了调整。以第一张为例正关键词:(masterpiece,highres,high quality,high resolution),(laba porridge:1.1),(red bean shrub:1.2),bowl,cured meat,bright light,take a high-angle/crane shot,chopsticks,red bean shrub,mung bean,black soya bean,red date,longan,lotus seed,chinese wolfberry,wooden table,put it on the table,shoot from top to bottom,photographic work,real photos,负面关键词NSFW,Negative,(worst quality:2),(low quality:2),(normal quality:2),watermark,ng_deepnegative_v1_75t,character,people,中文关键词:(杰作,高分辨率,高品质,高分辨率),腊八粥,红豆,碗,腊肉,明亮的光线,高角度拍摄,大米,小米,筷子,,绿豆,黑豆,木桌,放在桌子上,从上到下拍摄,摄影作品,ControlNet设置:预处理器:无模型:control_v1p_sd15_qrcode_monster_v2权重:1.1控制模型:均衡预处理器:depth_leres模型:control_v11f1p_sd15_depth权重0.3控制模型:更偏向提示词
关键词masterpiece,best quality,Photography,real photo,DOF,cute anthropomorphic cat,anthropomorphic,wearing a yellow translucent raincoat and rain boots,holding an umbrella,happy expression,masterpiece,full body,crazy details,Natural light,Bokehlicious,Fuitim X-T4Negative prompt:(worst quality,low quality:1.4),(malformed hands:1.4),(poorly drawn hands:1.4),(mutated fingers:1.4),(extra limbs:1.35),(poorly drawn face:1.4),missing legs,(extra legs:1.4),missing arms,extra arm,ugly,huge eyes,fat,worst face,(close shot:1.1),text,watermark,blurry eyes,设置Steps:38,Sampler:Euler a,CFG scale:7,Seed:1820151509,Size:512x512,Model hash:ed989d673d,Model:dreamshaper_7,Denoising strength:0.54,Clip skip:2,ENSD:31337,Hires upscale:2,Hires upscaler:Latent,Version:v1.5.1我们会得到一个修长动人的——猫。但当我们加上Cute Cat LoRA(下为0.6权重)后,小猫居然chibi起来了。LoRA不仅加入了很多可爱样子的猫,同时也把比例和画面调整。1.3.兔兔LoRA还没完,还有最后一个杀招!兔兔LoRA!太阔爱了,真的太阔耐了啊!!最主要的是,它可以让所有的猫猫狗狗都兔兔化,尤其是那个比例。三寸丁的身材。太阔爱了啊!!呜!~仍然是上面的关键词和参数,模型仍然是dreamshaper_7,什么都没调整,虽然这几个LoRA作者都建议使用chilloutmix_NiPrunedFp32Fix,这个随意。可以都试试的。
去秋叶的视频下载整合包[https://www.bilibili.com/video/BV17d4y1C73R/?spm_id_from=333.788.video.desc.click&vd_source=e99f85042059f2864f5cca20d71575f0](https://www.bilibili.com/video/BV17d4y1C73R/?spm_id_from=333.788.video.desc.click&vd_source=e99f85042059f2864f5cca20d71575f0),然后去另一个视频下载启动器,[https://www.bilibili.com/video/BV1ne4y1V7QU/?spm_id_from=333.1007.top_right_bar_window_history.content.click&vd_source=e99f85042059f2864f5cca20d71575f0](https://www.bilibili.com/video/BV1ne4y1V7QU/?spm_id_from=333.1007.top_right_bar_window_history.content.click&vd_source=e99f85042059f2864f5cca20d71575f0)把启动器解压到整合包根目录里启动。去civitai下载ChilloutMix[https://civitai.com/models/6424/chilloutmix](https://civitai.com/models/6424/chilloutmix)模型放到整合包models\Stable-diffusion目录里启动web UI切换模型。在web UI下载LoRA插件【Kohya-ss Additional Networks】并且启动。