开头致谢@秋葉aaaki 的懒人包合集以及LoRA训练脚本。如果你也有兴趣训练自己的LoRA模型的话，建议观看以下视频配置基础环境。训练要求显卡显存6GB以上。

模型发布地址

Civit

https://civitai.com/models/9895/reuve-starlight-anime-cg-style-lora-aruruotsuki

抱脸(huggingface)

https://huggingface.co/bdsqlsz/SD-LoRA-RevueStarlight-Anime-Style

模型分为原版和压缩版，压缩版大小为4MB，权重比原版低一些，使用需要增加0.2的权重值。压缩版0.8=原版0.6

模型介绍：

少女歌剧动漫游戏系列角色，大月阿露露LoRA模型。

特征为金发、长发、Smile系美少女，战斗服为牛仔服套装。

不是真人模型，请注意。动漫CG训练的，不过可配合真人模型使用。

ChilloutMix、Sunshinemix、Prefectworld等等

模型参数

训练图共32张，其中16组每张两组，由宽图(768X512)、长图(512X768)各一组组成，宽图为CG级人物图，含背景。长图为独立人物透明背景。重复次数10，批次20，训练共计6400步。

使用

作为关键字触发，在prompt中加入即可。

同时给相同姿势训练样图进行标记方便使用，可通过prompt

来触发指定样本图，配合使用找到需要的背景或姿势，修改token后的数字可更换样本标记，共1-16个训练样本可选。

基础模型：Yabuki_kentaro style(矢吹健太朗风格模型，2D使用推荐)

推荐权重：0.6

推荐采样器：DDIM(动漫或真人)，DPM++ SDE Karre(真人)

推荐VAE：kl-f8-anime2(动漫),vae-ft-mse-840000-ema-pruned(真人)

效果图

以上为测试效果图所加通用prompt，后续不再写出。

样例一：复活节兔女郎

promot：aruru_otsuki\(starlight\), aruru_otsuki\(starlight\)token4, transparent_background, 1girl, rabbit_ears, fur_trim, gloves, white_gloves, solo, animal_ears, bow, open_mouth, blonde_hair, smile, rabbit_tail, bowtie, long_hair,looking_at_viewer, yellow_eyes, sleeveless, bunny_pose, 

参数：Steps: 30, Sampler: DDIM, CFG scale: 11, Seed: 1712861818, Face restoration: CodeFormer, Size: 448x640, Model hash: fc2511737a, Model: chilloutmix_NiPrunedFp32Fix, Denoising strength: 0.7, Clip skip: 2, ENSD: 31337, Hires resize: 768x1024, Hires steps: 20, Hires upscaler: R-ESRGAN 4x+ Anime6B, AddNet Enabled: True, AddNet Module 1: LoRA, AddNet Model 1: aruru_otsuki(starlight)(e3f6f7f0f12b), AddNet Weight A 1: 0.6, AddNet Weight B 1: 0.6, Eta DDIM: 0.1

prompt同上。

Steps: 30, Sampler: DDIM, CFG scale: 11, Seed: 2935429010, Face restoration: CodeFormer, Size: 448x640, Model hash: a83ebdb42e, Model: yabukiKentaroStyle_yabukiFinalV10, Denoising strength: 0.7, Clip skip: 2, ENSD: 31337, Hires resize: 768x1024, Hires steps: 20, Hires upscaler: R-ESRGAN 4x+ Anime6B, AddNet Enabled: True, AddNet Module 1: LoRA, AddNet Model 1: aruru_otsuki(starlight)(e3f6f7f0f12b), AddNet Weight A 1: 0.6, AddNet Weight B 1: 0.6, Eta DDIM: 0.1

样例二：小红帽

prompt：aruru_otsuki\(starlight\), aruru_otsuki\(starlight\)token6, sparkle, tree, bush, nature, christmas_tree, flower, forest, 1girl, grass, hood, outdoors, diffraction_spikes, day, blonde_hair, basket, lens_flare, little_red_riding_hood_\(grimm\), solo, sky, palm_tree, light_rays, plant, white_flower, open_mouth, little_red_riding_hood_\(grimm\)_\(cosplay\), smile, glint, christmas_ornaments, pine_tree, lake, capelet, jewelry, bow, star_\(sky\), red_hood, hooded_capelet, frills, daisy, sunlight, 

参数：Steps: 30, Sampler: DDIM, CFG scale: 11, Seed: 663903066, Face restoration: CodeFormer, Size: 640x512, Model hash: fc2511737a, Model: chilloutmix_NiPrunedFp32Fix, Clip skip: 2, ENSD: 31337, Eta DDIM: 0.1, AddNet Enabled: True, AddNet Module 1: LoRA, AddNet Model 1: aruru_otsuki-000018(58271c43c88f), AddNet Weight A 1: 0.4, AddNet Weight B 1: 0.4

prompt同上

参数：Steps: 30, Sampler: DDIM, CFG scale: 7, Seed: 2929728208, Face restoration: CodeFormer, Size: 640x512, Model hash: a83ebdb42e, Model: yabukiKentaroStyle_yabukiFinalV10, Clip skip: 2, ENSD: 31337, AddNet Enabled: True, AddNet Module 1: LoRA, AddNet Model 1: aruru_otsuki-000012(e3f6f7f0f12b), AddNet Weight A 1: 0.6, AddNet Weight B 1: 0.6, Eta DDIM: 0.1

样例三：布伦希尔德(这图有点叩)

prompt：aruru_otsuki\(starlight\), 1girl, sparkle, fingerless_gloves, thighhighs, cloud, braid, weapon, sky, blue_sky, blonde_hair, long_hair, castle, gloves, holding_weapon, holding_sword, sword, earrings, holding, bow, dress, day,

参数：Steps: 28, Sampler: DPM++ SDE Karras, CFG scale: 11, Seed: 3905600500, Face restoration: CodeFormer, Size: 640x384, Model hash: 5cec660d69, Model: perfectWorld_perfectWorld, Denoising strength: 0.7, Clip skip: 2, ENSD: 31337, Hires upscale: 2, Hires steps: 20, Hires upscaler: ESRGAN_4x, AddNet Enabled: True, AddNet Module 1: LoRA, AddNet Model 1: aruru_otsuki(starlight)(e3f6f7f0f12b), AddNet Weight A 1: 0.6, AddNet Weight B 1: 0.6, AddNet Module 2: LoRA, AddNet Model 2: taiwanDollLikeness_v10(aa12327d4bab), AddNet Weight A 2: 0.4, AddNet Weight B 2: 0.4

prompt同上

参数：Steps: 28, Sampler: DDIM, CFG scale: 11, Seed: 345673101, Face restoration: CodeFormer, Size: 640x384, Model hash: f713bab753, Model: yesmix_v15, Denoising strength: 0.7, Clip skip: 2, ENSD: 31337, Hires upscale: 2, Hires steps: 20, Hires upscaler: R-ESRGAN 4x+ Anime6B, AddNet Enabled: True, AddNet Module 1: LoRA, AddNet Model 1: aruru_otsuki(starlight)(e3f6f7f0f12b), AddNet Weight A 1: 0.6, AddNet Weight B 1: 0.6, Eta DDIM: 0.1

以上，其他请大家自己去尝试一下。

近期会做一个LoRA训练视频教程，敬请期待。