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 pixivID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 amchosen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 bothchosen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.