<h1 id="niji-noob-37sie6k2q">Niji Noob</h1><a target="_blank" rel="ugc" href="https://civarchive.com/models/176554/usnr-style">Carried by USNR STYLE</a>NoobAI XL v-pred model (requires different settings than usual to get the most of it, see below)Merged by dumping web UI model from memory to diskMy contribution: I've balanced the painterly look with coherence and used some of my own LoRA's for adjustment. Settings:<pre><code>Steps: 14, Sampler: DPM++ SDE CFG++, Schedule type: SGM Uniform, CFG scale: 1.7, Size: 832x1216</code></pre>Tool: <a target="_blank" rel="ugc" href="https://github.com/Panchovix/stable-diffusion-webui-reForge">Panchovix/stable-diffusion-webui-reForge</a>Tips:<ul><li>This is a v-pred model so CFG needs to be very low depending on the sampler/scheduler or you will fry your gens.</li><li>Lower CFG (e.g. 1.1) will produce less saturated results. I like my gens saturated so I use 1.7.</li><li>You really only need 15 steps at most, less for hiresfix.</li><li>reForge has the best memory management of all the SD webUI's so I can output 1664 x 2432 no worries on a GTX 3090 (using 2x on hiresfix).</li><li>Use NoobAI artist tags if you don't like / want to adjust the default look.</li><li>Non v-pred (as in normal) LoRAs works fine (tbh even better in most cases).</li><li>If you get tired of circles in backgrounds the solution is simple: put "tunnel" in your neg prompt.</li><li>For more Niji-like output use simple prompts and let the model steer your gens</li><li>Want to merge your own similar model, check out the LoRA's used by exlight: <a target="_blank" rel="ugc" href="https://aibooru.online/posts/122961?q=exlight">AIBooru</a></li><li>Also check out LyiaEngine for CivitAI backups of the above: <a target="_blank" rel="ugc" href="https://huggingface.co/LyliaEngine">LyliaEngine (Narontaka)</a></li></ul>

Files with SHA256: ad8cf12367632d8e4edd7a56827f58f76dc25a7ad553ba5cb866351a95a204ca