这个lora是重新上传的，例图重新更新。V5 V7 V8 V9是之前的版本。V10是本次新更新的版本This LoRa version has been re-uploaded, and the example images have been updated. V5, V7, V8, and V9 are previous versions. V10 is the newly updated version.<h2 id="cunnyfunkylikefollow"><a target="_blank" rel="ugc" href="https://www.pixiv.net/users/113801960">Cunnyfunky</a>⬅❤Like➕Follow</h2>建议权重：0.7~1.0 Recommended weight: 0.7~1.0 图幅大小建议：1024x1344 1024x1408 1024x1536 Suggested map sizes:：1024x1344 1024x1408 1024x1536<hr /><h2 id="lora">本人制作的所有lora禁止用于商业用途！！！</h2><h2>特此说明！！！</h2><h2 id="all-loras-i-created-are-prohibited-from-being-used-for-commercial-purposes!!!">All loras I created are prohibited from being used for commercial purposes!!!</h2><h2 id="this-is-to-clarify!!!">This is to clarify!!!</h2><hr /><h3 id="v10lorav10">V10是我全新融合的版本。 这个lora的皮肤是哑光质感的皮肤，脸部和眼镜也非常清晰。没有灰色的雾气。为了庆祝重新上传，所以制作了V10这个版本</h3><h3 id="v10-is-my-newly-merged-version.-this-lora-skin-has-a-matte-finish-and-the-face-and-glasses-are-very-sharp.-there's-no-gray-haze.-i-created-this-v10-version-to-celebrate-the-re-upload."> V10 is my newly merged version. This Lora skin has a matte finish, and the face and glasses are very sharp. There's no gray haze. I created this V10 version to celebrate the re-upload.</h3><hr />

Cunnystyle

CunnyfunkyV7.0.safetensors

v7.0

v10.0

v9.0

v8.0

v5.0

Weight: 0.5~0.7特别鸣谢 一个友人：阿荣 刚开始训练V1.0是有他的推荐才开始了第一次的训练，经过了多轮的迭代尝试，画风定格到了5.0。在此过程中每每挑选素材，阿荣都会和我一起挑选。<h2 id="cunnyfunkylikefollow"><a target="_blank" rel="ugc" href="https://www.pixiv.net/users/113801960">CunnyFunky</a>⬅❤Like➕Follow</h2>Sampler：Euler aSteps：30 CFG：5 1024*1536 1024*1344 896*1536<hr />放大算法建议使用：4x-AnimeSharp 迭代步数：20 重绘幅度：0.09 CFG：8 AD修脸 重绘幅度0.2~0.4，使用独立宽高，1024*1024Hires.fix: 4x-AnimeSharpsteps: 20Redraw amplitude: 0.09 CFG：8<hr />AD face trim, redraw amplitude 0.4,Use independent width and height（1024 * 1024）<hr /><h2 id="lora">本人制作的所有lora禁止用于商业用途！！！</h2><h2>特此说明！！！</h2><h2 id="all-loras-i-created-are-prohibited-from-being-used-for-commercial-purposes!!!">All loras I created are prohibited from being used for commercial purposes!!!</h2><h2 id="this-is-to-clarify!!!">This is to clarify!!!</h2><hr /><h2 id="v9.0">V9.0</h2>V9.0采用了cunnyfunky最新的素材，添加了正则的使用。肢体会有更好的表现，并且消除了人物脸部灰色朦胧的滤镜，脸部更加清晰，表情更加完美。这一版本中的屁股表现尤其出色。V9.0 uses the latest material from cunnyfunky and adds the use of regularity. The limbs will perform better, and the gray and hazy filter of the character's face will be eliminated, resulting in a clearer face and a more perfect expression. The butt performance in this version is particularly good.<h2 id="v8.0">V8.0</h2>V8.0回归源初，虽然7.0眼睛和脸部是最清晰的，但是偏离了Cunny原本的味道。V8重新寻找了Cunny新的素材。重新炼制，迭代了4个小版本，目前的效果是我最喜欢的。V8.0 returns to the beginning, although 7.0 eyes and face is the most clear, but deviated from the original taste of Cunny. V8 re-found Cunny new material. Rerefined, iterated 4 small versions, the current effect is my favorite.<h2 id="v7.0">V7.0</h2>V7.0生成的脸以及眼睛和cunny的完全不同。但是，这个版本的眼睛和脸是最清晰的。V7.0 is my most satisfactory version of Cunny, although the generated face and eyes are completely different from cunny's. However, this version has the clearest eyes and face.<h2 id="v5.0">V5.0</h2>V5.0是我个人觉得最像Cunny的一个版本，也是脸部崩坏不严重的版本。很好地还原了脸和眼睛，还有极大的保证了手不会崩坏。V5.0 is the version that I personally think is the most like Cunny, and it is also the version with less serious face collapse. It restores the face and eyes well, and greatly guarantees that the hands will not collapse.<h2 id="v4.0">V4.0</h2>V4.0我更换了所有的训练集，根据以往的经验重新炼制了这个版本。他的缺点是眼睛和脸不那么完美，手部有时候也会崩坏，但是这确实是我的一次炼丹水平的提高。V4.0 I replaced all the training datasets and rerefined this version based on past experience. Its shortcomings are that the eyes and face are not perfect, and the hands sometimes collapse, but this is indeed an improvement in my alchemy level.V1.0~3.0V1.0是差异化炼丹，V2.0和V3.0都是对Cunny的尝试，这些版本为较早期的版本。V1.0 is differentiated alchemy, V2.0 and V3.0 are both attempts at Cunny, and these versions are earlier versions.

cunnystyle

V6.0

V5.0

v4.0

v3.0

v2.0

v1.0

<a target="_blank" rel="noopener noreferrer nofollow" href="https://civitai.com/models/1506082/cunnystyle">cunnystyle - v7.0 | Illustrious LoRA | Civitai</a>Weight: 0.5~1.0<a target="_blank" rel="ugc" href="https://www.pixiv.net/users/113801960">CunnyFunky</a>⬅❤Like➕FollowV6.0I recalibrated the cunny tone and blended it with the lora and brightsky I made earlier, changing the face shape of the cunny's original lines, making the eyes more complete, the tone saturation is higher, and there is some matte skin texture.我重新调整了cunny的色调，并且把它和我之前做的一个lora，brightsky相融合了，改变了cunny原有线条的脸型，让眼部更加完整，色调饱和度更高了一些，并且有一些哑光的皮肤质感。V5.0I added some close-up material base images of faces, and additional base images of limbs, so that the model can have richer body movements.我增加了一些脸部特写的素材底图，额外增加了一些肢体的底图，让模型更加有更加丰富的体态动作。V4.0In version 4.0, I replaced all the footage again and adjusted the learning rate. The painting style and limbs have been greatly improved, and the tone is closer to the effect of Cunny itself.4.0版本 我再次更替了所有的素材图，又调整了学习率。使得画风和肢体有了非常大的改善，色调也更加的接近Cunny本身的效果。V3.0Version 3.0 I replaced all the material images to make the function and effect of this Lora more comprehensive, especially for the effect of the face, which is more inclined to cunny. This Lora should be the last version.3.0版本 我更替了所有的素材图，使这个Lora的功能效果更加的全面，尤其对于脸部的效果，更加偏向于cunny。这个lora应该是最后一版了。V2.0Version 2.0 is an alchemy of all material pictures. It will change the structure of the original model, but the resulting picture will be more like the effect of cunny.2.0版本是全素材图的炼丹，它会改变原模型的结构，但是出来的图会更像cunny的效果。V1.0 Version 1.0 is a differentiated Lora, which makes minimal changes to the image and native model output composition.1.0版本是差异化的lora，他对于图像和原生模型出图的构图改动是最小的

Weight: 0.5~1.0

CunnyFunky⬅❤Like➕Follow
V6.0
I recalibrated the cunny tone and blended it with the lora and brightsky I made earlier, changing the face shape of the cunny's original lines, making the eyes more complete, the tone saturation is higher, and there is some matte skin texture.
我重新调整了cunny的色调，并且把它和我之前做的一个lora，brightsky相融合了，改变了cunny原有线条的脸型，让眼部更加完整，色调饱和度更高了一些，并且有一些哑光的皮肤质感。
V5.0
I added some close-up material base images of faces, and additional base images of limbs, so that the model can have richer body movements.
我增加了一些脸部特写的素材底图，额外增加了一些肢体的底图，让模型更加有更加丰富的体态动作。
V4.0
In version 4.0, I replaced all the footage again and adjusted the learning rate. The painting style and limbs have been greatly improved, and the tone is closer to the effect of Cunny itself.
4.0版本 我再次更替了所有的素材图，又调整了学习率。使得画风和肢体有了非常大的改善，色调也更加的接近Cunny本身的效果。
V3.0
Version 3.0 I replaced all the material images to make the function and effect of this Lora more comprehensive, especially for the effect of the face, which is more inclined to cunny. This Lora should be the last version.
3.0版本 我更替了所有的素材图，使这个Lora的功能效果更加的全面，尤其对于脸部的效果，更加偏向于cunny
。这个lora应该是最后一版了。
V2.0
Version 2.0 is an alchemy of all material pictures. It will change the structure of the original model, but the resulting picture will be more like the effect of cunny.
2.0版本是全素材图的炼丹，它会改变原模型的结构，但是出来的图会更像cunny的效果。
V1.0
Version 1.0 is a differentiated Lora, which makes minimal changes to the image and native model output composition.
1.0版本是差异化的lora，他对于图像和原生模型出图的构图改动是最小的

Cunny v4

Cunnystyle | Illustrious

Files with SHA256: 663d5724863ae01c4566ee7d2f8c7218d03a5b53c601198ed9fc21ebb4f50960