⚠️ This model is my first one and is in alpha. There is some known issues and bias that need to be fixed.This model tries to enhance dynamic anime-like effects; such as <a target="_blank" rel="ugc" href="https://danbooru.donmai.us/wiki_pages/motion_lines">motion lines</a>, <a target="_blank" rel="ugc" href="https://danbooru.donmai.us/wiki_pages/emphasis_lines">emphasis lines</a>, <a target="_blank" rel="ugc" href="https://safebooru.donmai.us/wiki_pages/speed_lines">speed lines</a>, <a target="_blank" rel="ugc" href="https://danbooru.donmai.us/wiki_pages/afterimage">afterimage, </a><a target="_blank" rel="ugc" href="https://danbooru.donmai.us/posts?tags=sound_effects">sound effects, </a>etc.It is meant to work with a illustrative style and don't work well with 3D or realistic images. I recommend any Pony models (checkpoints, additional LoRAs) with an anime-based style.Aside the triggering words, you can use some of the following tags to enhance the dynamic.<pre><code>dynamic lines, motion lines, speed lines, afterimage, emphasis lines, sound effects, speech bubble, spoken heart</code></pre>For recommended parameters and additional information, check Usage Tips and About this version.

Dynamic Lines

This version don't really improve the v0.2, but it is trained to have a drawing style. It improves lineart, sketch and other doujinshi-like render.You can experiment by adding tags like <code>doujinshi, lineart, pastel colors</code> in your prompt and <code>3d, realistic</code> in the negatives.

Files with SHA256: 8cc7e720a4870de3f7cca74779efec7bce242f8d29e051dc038999abd9c6b1d6