CivArchive
    水晶球/Crystal ball - v1.0
    NSFW
    Preview 5411920
    Preview 5411915
    Preview 5411899
    Preview 5411897
    Preview 5411910
    Preview 5411896
    Preview 5411913
    Preview 5411911
    Preview 5411912
    Preview 5411901
    Preview 5411906
    Preview 5411914
    Preview 5411918
    Preview 5411919
    Preview 5411916
    Preview 5411917
    Preview 5411974

    大家好我是chosen鲷,这次给大家带来水晶球lora。

    顾名思义,就是如同例图中那种感觉的,水晶球里有各种人物景物之类的,小动物也可以,并不局限于人物,我训练集里有各种内容。效果还是挺唯美的,但是由于人物占比太小,所以比较容易崩手崩脸,建议加上修脸插件使用。

    然后顺便说一下,tag尽量简介,tag写多了它就出不来水晶球了,lora本身很好触发,都不需要触发词。

    lora兼容性还可以,建议权重0.8。

    推荐的tag:

    (masterpiece:1.2),best quality,PIXIV,

    这里写你想要的内容

    <lora:Crystal ball_20240112152706:0.8>,
    Negative prompt: EasyNegative,badhandsv5-neg,Subtitles,word,
    Steps: 32, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 3978777272, Size: 512x768, Model hash: e3edb8a26f, Model: ghostmix_v20Bakedvae, Denoising strength: 0.5, Clip skip: 2, ADetailer model: face_yolov8n.pt, ADetailer confidence: 0.3, ADetailer dilate erode: 4, ADetailer mask blur: 4, ADetailer denoising strength: 0.4, ADetailer inpaint only masked: True, ADetailer inpaint padding: 32, ADetailer version: 23.11.0, Hires upscale: 2, Hires steps: 12, Hires upscaler: R-ESRGAN 4x+, Lora hashes: "Crystal ball_20240112152706: c161e0195d75", TI hashes: "EasyNegative: 66a7279a88dd, badhandsv5-neg: aa7651be154c", Version: v1.6.0-2-g4afaaf8a

    大模型我推荐用ghost加上YOZORA的vae以达到最佳效果(下方推荐),其他的常见常见二次元模型也均可。

    感谢大家的下载使用,如果喜欢的话欢迎关注我,我的bilibili pixiv ID均为chosen鲷,欢迎大家关注我的频道https://space.bilibili.com/3867754?spm_id_from=333.1007.0.0,https://www.pixiv.net/users/40911695 谢谢。

    The following content is from Google Translate

    Hello everyone, I am chosen鲷. This time I bring you Crystal Ball Lora.

    As the name suggests, it's like the feeling in the example image, with various characters and scenes inside the crystal ball. It can include small animals as well, not limited to just human characters. My training dataset contains various content. The effect is quite beautiful, but because the proportion of characters is small, it is more prone to blurring. I recommend using a face retouching plugin.

    By the way, please keep the tags brief. If there are too many tags, the crystal ball effect may not appear. Lora itself is easy to trigger, and no specific trigger word is needed.

    Lora has good compatibility, and I recommend using weight 0.8.

    Recommended tags:

    (masterpiece:1.2), best quality, PIXIV,

    Write here the content you want

    <lora:Crystal ball_20240112152706:0.8>, Negative prompt: EasyNegative, badhandsv5-neg, Subtitles, word, Steps: 32, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 3978777272, Size: 512x768, Model hash: e3edb8a26f, Model: ghostmix_v20Bakedvae, Denoising strength: 0.5, Clip skip: 2, ADetailer model: face_yolov8n.pt, ADetailer confidence: 0.3, ADetailer dilate erode: 4, ADetailer mask blur: 4, ADetailer denoising strength: 0.4, ADetailer inpaint only masked: True, ADetailer inpaint padding: 32, ADetailer version: 23.11.0, Hires upscale: 2, Hires steps: 12, Hires upscaler: R-ESRGAN 4x+, Lora hashes: "Crystal ball_20240112152706: c161e0195d75", TI hashes: "EasyNegative: 66a7279a88dd, badhandsv5-neg: aa7651be154c", Version: v1.6.0-2-g4afaaf8a

    For the large model, I recommend using ghost together with YOZORA's VAE for the best results (recommended below). Other common anime-style models are also suitable.

    Thank you all for downloading and using. If you like it, please follow me. My bilibili pixiv IDs are both chosen鲷. Welcome to follow my channels on Bilibili https://space.bilibili.com/3867754?spm_id_from=333.1007.0.0 and Pixiv https://www.pixiv.net/users/40911695. Thank you.

    Description

    FAQ

    Comments (9)

    GoldenMiocolaJan 12, 2024
    CivitAI

    这个很棒捏💕

    chosen
    Author
    Jan 13, 2024· 1 reaction

    嗯,效果很唯美

    itachiiiJan 14, 2024
    CivitAI

    you are the best lora creator sir and that's why i choose you to train the 90's anime style midjourney style lora

    == can you please train the lora ==

    here is some examples -

    https://cdn.discordapp.com/attachments/997267880276136037/1194397618420318328/kymerasow_90s_anime_style_girl_c56bd4fa-acb8-44ad-b6e1-adc0717bccd0.png?ex=65b03477&is=659dbf77&hm=3d6f5401cf4cb38b166aefad3ba5370bf8b66e1efe004c393e21daa77ad187e0&

    https://cdn.discordapp.com/attachments/1103792881736630395/1194398150509740042/kymerasow_90s_anime_style_girl_1d147b82-af99-4bef-94b5-45300a0a9000.png?ex=65b034f5&is=659dbff5&hm=a9b3747449a9ad57efdd5b5d28fd552ca5e149339880f6f780b9c9b3883a80f3&

    https://cdn.discordapp.com/attachments/1103792881736630395/1194398154188140674/kymerasow_90s_anime_style_girl_fa820e11-3c1a-4771-8820-704f4efb2e81.png?ex=65b034f6&is=659dbff6&hm=1435bb3353fefa7395716ef84452657b485b528913ab1e1a9ab303cf88b0af89&

    https://cdn.discordapp.com/attachments/1103792881736630395/1194398148349657128/kymerasow_90s_anime_style_girl_f954db5b-aeee-4e3d-a23c-b657d8f8142f.png?ex=65b034f5&is=659dbff5&hm=c22bd183b75bc5d7b13a4bf84dac47f979f9cc1cab342f04d4a523aea34bba0e&

    https://cdn.discordapp.com/attachments/997267880276136037/1194397641006657566/kymerasow_90s_anime_style_girl_9d956f45-d082-4fe5-bc56-d60b47376f0a.png?ex=65b0347c&is=659dbf7c&hm=06dc3d160516218da858ba7fe28a7607f5bb4444ae8bedb402ec5de4764249cf&

    https://cdn.discordapp.com/attachments/997267880276136037/1194396444933759006/kymerasow_90s_anime_style_girl_8a1fa07c-cbf2-4f9a-9ac0-9a8cec48b2ef.png?ex=65b0335f&is=659dbe5f&hm=5ce4d643a8822855ceb467bd21301e1df3ed0002f663b05dd224cc95d2ea49e8&

    https://cdn.discordapp.com/attachments/997267880276136037/1194395158561689690/kymerasow_90s_anime_style_girl_9e93fc36-8143-4cfa-b6da-d75b9d087d5d.png?ex=65b0322c&is=659dbd2c&hm=db6ac9f01dd4aa777cd4248a1a91e506d1c5f3426d4feadcf67d76641bdaba6c&

    chosen
    Author
    Jan 15, 2024

    I've already trained this style of lora. Here's the link https://civitai.com/models/124213/981990s-animation-style-lora?modelVersionId=135566

    itachiiiJan 16, 2024

    @chosen thank you so much

    but what about background?

    chosen
    Author
    Jan 16, 2024

    @itachiii I haven't done any background training because there is no background training set of this type

    itachiiiJan 16, 2024

    @chosen 

    https://civitai.com/models/124213/981990s-animation-style-lora?modelVersionId=135566

    but you don't trained this on midjourney V6

    please train on V6 and update

    itachiiiJan 16, 2024

    @chosen i wish i know the lora training

    nobody55Jan 29, 2024
    CivitAI

    把美少女装球里需要几步

    LORA
    SD 1.5

    Details

    Downloads
    516
    Platform
    CivitAI
    Platform Status
    Available
    Created
    1/12/2024
    Updated
    5/13/2026
    Deleted
    -
    Trigger Words:
    Crystal ball

    Files

    Crystal ball_20240112152706.safetensors

    Mirrors

    Available On (1 platform)

    Same model published on other platforms. May have additional downloads or version variants.