<h3 id="introduction-na746es3v">Introduction</h3>Fluffy Tart ILXL is a model that aims to generate heavily styled images. It was made for generating furry characters, but human characters works as well.This model was made by merging <a target="_blank" rel="ugc" href="https://civarchive.com/models/1358997?modelVersionId=1535163">https://civarchive.com/models/1358997?modelVersionId=1535163</a> as the base model and a custom made LoRA.<h3 id="recommended-settings-mgiy3j0xf">Recommended Settings</h3><ul><li>CFG : 1 ~ 3</li><li>Steps : 25 ~ 30</li><li>Sampler : ddim</li><li>Size : 768 * 1024(I mainly use this size)</li></ul><h3 id="starter-prompt-od2ztryqq">Starter Prompt</h3><pre><code>masterpiece, best quality, 1girl, solo,</code></pre><pre><code>worst quality, bad quality, off-topic, username, logo, signature, watermark, text,</code></pre><h3 id="warning-6zcicp0ic">Warning</h3>Model loves to generate small background with white frame. Consider using tags like <code>outdoors</code> , <code>indoors</code>.

Fluffy Tart ILXL [NoobAI vpred]

fluffyTartILXLNoobai_apple.safetensors

Apple

This is NOT my model. Made by <a rel="ugc noopener noreferrer nofollow" href="https://civitai.com/user/delicioustart" target="_blank">delicioustart</a> on Civitai, this is just a reprint.If you're the creator of this model and want this reprint taken down, leave a comment below and I will do so.Prompting:<ul><li>Quality Tags: masterpiece, best quality,</li><li>Negative Tags: worst quality, bad quality, lowres, bad hands, extra digits, fewer digits, off-topic, text, signature, watermark, logo,</li><li>Trigger words are optional</li><li>Your positive prompt should be structured like this from front to back:quality &amp; aesthetic tags, artist (s), 1girl/1boy, character, series, general tags</li><li>Positive prompt example: "masterpiece, best quality, 1girl, frieren, sousou no frieren, solo, white robe, twintails, black pantyhose, looking at viewer, angry, hands on own hips, cowboy shot, dutch angle, outdoors, nature, wind"</li><li>Use Danbooru tags to word your prompts -&gt; <a rel="ugc noopener noreferrer nofollow" href="https://danbooru.donmai.us/wiki_pages/tag_groups" target="_blank">Danbooru Tag Groups Wiki</a></li><li>When adding artist tags, do not use "by" or "artist:", simply put the artist name</li><li>Do not use underscores anywhere in your prompt unless they are for a LoRA trigger word</li></ul>Settings:<ul><li>Automatic VAE</li><li>Use SDXL resolutions: 1024x1024, 896x1152, 832x1216, 768x1344, 640x1536</li><li>Euler a, DDIM or DPM++ 2M Karras</li><li>25-30 steps</li><li>1-6 CFG</li><li>Clip Skip 2</li></ul>

Files with SHA256: 97ccdaa4900cad01ccbb7e4fc718e239405c372619816b9e02d5a35dad86b8d4