<ul><li>without hires to gen highly detailed photo</li><li>without default asian face</li><li>without need too much tag weight, better text understanding</li><li>works fine with almost all LORA and TI</li></ul><h3 id="heading-173">[update:0230825]</h3><ul><li>version 2 (close to sdxl1.0)</li></ul>Advices<ol><li>Not require strong prompt weight adjustments. Try removing all prompt weights.</li><li>Not require Hi-RES. Try big steps (&gt;50steps) and use a DPM++ series sampler.</li><li>Asian-style version 👉 <a target="_blank" rel="ugc" href="https://civarchive.com/models/9052?modelVersionId=146253">LOFI-v4</a></li><li>Fantasy-style version 👉 <a target="_blank" rel="ugc" href="https://civarchive.com/models/133257/epic-la">EPIC-v1</a></li></ol><h3 id="heading-178">[update: v1]</h3><h1 id="heading-298">What raw-xs ?</h1><ul><li>raw: Realistic Artistic Wizardry</li><li>xs: Extra Small</li></ul><s>(chatgpt generated)</s><s>a little better than sdxl-0.9 （XD</s>based on <a target="_blank" rel="ugc" href="https://civarchive.com/models/9052?modelVersionId=102152">LOFI_V3</a> (with sdxl0.9) fine-tuning<h3 id="heading-300">Tips</h3>this model may be include wrong sdxl partten 😅(for example, in some color modes, human skin will be very oily, You can try different samplers, there will be different effects)<h3 id="heading-301">Backup</h3><a target="_blank" rel="ugc" href="https://huggingface.co/lenML/raw-xs/tree/main">https://huggingface.co/lenML/raw-xs/tree/main</a>

Files with SHA256: b590da8b7fa213fa16f15c695d093a8ad13386f308d9ca1fe06e1879911795d7