<p>These loras were trained on images from <a target="_blank" rel="ugc" href="https://danbooru.donmai.us/wiki_pages/megami_magazine">Megami Magazine</a> from late 2018 to 2026. Anima already knows the Megami Magazine tag but it's riddled with japanese text. These loras were trained with the text manually removed. Please read the <strong>About This Version</strong> for more information.</p><p>To get the results you see in the images above, <a target="_blank" rel="ugc" href="https://civarchive.com/articles/11427/my-current-sdxl-settings">these are the settings I used.</a></p><p><span style="color:rgb(250, 176, 5)">Visit My Site For Models I Can't Post On CivitAI:</span> <a target="_blank" rel="ugc" href="https://qtgurl.com/">https://qtgurl.com/</a></p><p><span style="color:rgb(18, 184, 134)">The loras for Anima V1 were trained at 1792 resolution. So they're capable of generating at around 1792x1792 or higher.(I use 1600x1984) The P2 and P3 loras were trained at 1536 resolution, so they're capable of generating around 1536x1536. (I use 1408x1792 for most gallery images.) All of the loras are capable of generating smaller sizes just fine.</span></p><p></p><h3 id="memaani-v2+b+c">MeMaAni - V2+B+C</h3><p>Trained on 48-68 (mostly) new and<strong> </strong>old images. Dataset was re-collected and edited for the best possible quality.</p><h3 id="memaani-v1">MeMaAni - V1</h3><p>Trained on the same data used in Preview 3.</p><h3 id="memaani-anima-preview-3">MeMaAni - Anima Preview 3</h3><p>Trained on the same data used in Preview 2 with slight changes to improve results. Most of the improvements come from the Anima P3.</p><h3 id="memaani-anima-preview-2">MeMaAni - Anima Preview 2</h3><p>Trained on the same data used in 2d Gold Fish. I guess you could call it MeMa F, but I suppose that would make it A for this model. It uses a lot of images from the A dataset so it's close enough.</p><p></p><h3 id="comment-on-anima">Comment On Anima</h3><p>The final release of Anima 1.0 is fantastic. The model has so much capability. If you haven't made the switch, now is the time. It doesn't matter if all you post is one girl, the jump in quality, prompt comprehension, 16 channel VAE, newer anime knowledge and just raw stable performance makes it the worth while model to switch to as an anime genner.</p><p></p>

MeMaAni Flat Anime Style - Anima

<p>This version was trained on 48 images at 1792 from the brand new C dataset. Most of the images are from the original MeMaXL C dataset with 20 additional images gathered to further flesh out the style.</p><p>The focus of this version, like the original version, was to reproduce eyes that were more slanted with round pupils. Basically, eyes similar to <a target="_blank" rel="ugc" href="https://danbooru.donmai.us/posts?tags=tsurime&amp;z=1">Tsurime</a>. The choice of images used had an added effect of vivid colors and random shading gradation. Lastly, just like the version before this, there was no emphasis put on backgrounds.</p>

Files with SHA256: c67628076e7a5228480779764823c136e002d38e71824cb516a18093815862ff