<h3 id="realand3dandanime-is-multi-style-base-models-not-limited-to-realism-alone.">Real&amp;3D&amp;Anime is multi-style base models, not limited to realism alone.</h3><pre><code>I don't want to destroy the anime concept of Anima. So i finetune this model in anime style by default.I'm aimming to use high-quality real cosplay and scenery photo data to achieve realistic style, and improve anime concept's anatomical\scene composition, through its realistic concept.</code></pre><h3 id="recommended-settings">Recommended settings：</h3><ul><li>postive promptsmasterpiece,best quality,newest,high resolution,</li></ul><ul><li>postive prompts for realisticphoto \(medium\) ,[ <code>photo \(medium\)</code> : 0.5–3]</li></ul><ul><li>negative prompts</li><li>worst quality,low quality,lowres,signature,text,logo,european,mature female,lowres,e621,fat,saegyo sal,puffy cheeks,thick lips,</li></ul><ul><li>Suggest parametersSampling Method：ER SDE / Eulr / Heun / DPM++ 2M SDESchedule Type：beta / simple cfg : 3~5（I often use 4.5 for realistic） shift：3 </li></ul><ul><li>Artist Style@ danbooru artist name</li></ul><h3 id="part-of-specially-trained-tags">Part of Specially Trained Tags</h3>Background: blurry foreground, blurry background (trained to help the model distinguish blurred areas)Art Style: photo (medium), 3D, impasto, draft, flat color, no lineart, ···Lighting: dark, chiaroscuro, night, dark background, backlighting, overexposure, ··· (Chiaroscuro delivers effects similar to camera flash; alternative tags are used in training to prevent the model from generating cameras)Composition: partially underwater shot, partially underground shot, partially submerged, dutch angle, pov crotch, female pov, foreshortening, ···Age: ···, mature female, old man, old woman, height difference, ···Ethnicity &amp; Creature: European, mermaid, monster, ···Visual Effects: bloom, vignetting, motion blur, motion line, chromatic aberration, ···Special Features: aegyo sal, puffy cheeks, entry plug (Aegyo sal refers to a popular eye makeup style widely seen in cosplay photos)<h3 id="the-following-tags-are-not-recommended-for-use:">The following tags are not recommended for use:</h3><ol><li>solo: Removed from training datasets; its use results in 2D outputs. Replace with solo focus instead.</li><li>Duplicate standalone tags such as "shirt, red shirt": Synonymous duplicate tags have already been filtered out to improve prompt description accuracy.</li></ol><h3 id="training-tool">Training Tool</h3>I use (<a target="_blank" rel="ugc" href="https://github.com/wochenlong/lora-scripts-next">https://github.com/wochenlong/lora-scripts-next</a>) for Full-finetuning. The tool is made by Chenkin (且听风吟，<a target="_blank" rel="ugc" href="https://civarchive.com/user/Chenkin">https://civarchive.com/user/Chenkin</a>).And thanks for his technical help.

RC-AIO 2.0B | Real、Anime、3d、figure Full-finetune

Trained on 14,000 images:<ul><li>40% Photo</li><li>30% Anime</li><li>15% 3D/Figure</li><li>15% Other</li></ul>

Files with SHA256: 1da16524800093aeb1a06e65cccd6389732f1d0b086609d37681cec5886e9bb9