All pics used the extension adetailer for the face.In negatives I can recommend the following. It's making a nice style: <pre><code>(loli, greyscale, monochrome:0.5)</code></pre>I made a LoRA for PonyXL using my old generated images with my sd1.5 models ObsidianV3-Flat and ObsidianV3.3.Can make some good abstract and colorful stuff.I also created a model called <a target="_blank" rel="ugc" href="https://civarchive.com/models/436819">ObsidianPDXL </a>using this lora. You can also support me on <a target="_blank" rel="ugc" href="https://www.patreon.com/NoctAI">Patreon</a> or <a target="_blank" rel="ugc" href="https://www.deviantart.com/noctfuta">DeviantArt</a>

NoctFlatStyle

Same dataset as V2 but I use XL base model

Files with SHA256: 70bd842829d0d4efeb012ff646bc572d14ed3639fd84057766d68bc2d58b79b5