Here is the LoRA based on Spikedmauler's style.The training data is divided into 3 general categories: sketch, drawing, paintingFor the best of 3 worlds, I would use the following: Positive<pre><code>score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, &lt;lora:StS_detail_slider_v1:1.3&gt;, &lt;lora:Spikedmauler_Style_Pony_XL:0.85&gt;, Sp1kedMauler, by spikedmauler, BREAK</code></pre> Negative<pre><code>{patreon, text, logo, signature},</code></pre> I am incredibly biased toward Spikedmauler's painterly rendering, so I recommend this: Positive<pre><code>warm lighting, cool shadows, bounce lighting, detailed background, BREAK realistic, subsurface scattering, cast shadow, painting:1.2, texture, score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, &lt;lora:StS_detail_slider_v1:1.3&gt;, &lt;lora:Spikedmauler_Style_Pony_XL:0.9&gt;, Sp1kedMauler, by spikedmauler, BREAK </code></pre> Negative<pre><code>{patreon, text, logo, signature}, simple background, [[sketch, drawing]], </code></pre>

Files with SHA256: 2c0d365ebd0d027fcbafc822c4374edff964a94ae75f91307317333cc54aa4fc