🎞 LoRA is trained purely on anime screencaps<hr />🗝 Activation tags are in the right panel📸 Always use ADetailer to get detailed eyes🛠 Technical notes:<ul><li>LoRA shoud be compatible with Illu models &amp; NoobAI. (Not tested.)</li><li>LoRA is compatible with various styles - Lower the weight &amp; remove "anime screencap" from the prompt, if you are getting too strong anime style</li><li>Hi. res fix used on all sample images:<ul><li>Upscaler: R-ESRGAN 4x+ Anime6B</li><li>Steps: 10-15</li><li>Denoising: 0.3</li></ul></li></ul><hr />📬Leave a like, comment, review but first of all, ENJOY ;)

Karyl [2 outfits] | Illustrious | Princess Connect! Re:Dive

<ul><li>fixed oversaturation / bloom</li><li>fixed washed out colors</li><li>removed bad images from the dataset</li><li>upscaled dataset images &amp; got rid of blurriness</li></ul>

PrincessConnect_Karyl_IlluXL.safetensors

Illu v1.1

Illu v1.0

&lt;ul&gt;&lt;li&gt;&lt;p&gt;fixed oversaturation / bloom&lt;/p&gt;&lt;/li&gt;&lt;li&gt;&lt;p&gt;fixed washed out colors&lt;/p&gt;&lt;/li&gt;&lt;li&gt;&lt;p&gt;removed bad images from the dataset&lt;/p&gt;&lt;/li&gt;&lt;li&gt;&lt;p&gt;upscaled dataset images &amp;amp; got rid of blurriness&lt;/p&gt;&lt;/li&gt;&lt;/ul&gt;

Files with SHA256: b1e6a4238ad71977882d639b918d719b3e6d04eb29c1bb751590f384428fbef7