Best settings:<ol><li>Sampling Method: Euler A or Euler A SGMUniform - you can mess around with this</li><li>Sampling Steps: 20 ~ 60 depending on your sampling method</li><li>Hires. fix: Enabled</li><li>Hires Steps: 15 ~ 30</li><li>Upscaler: <a target="_blank" rel="ugc" href="https://openmodeldb.info/models/4x-NMKD-YandereNeo-XL">4x_NMKD-YandereNeoXL_200k</a> - You can test others.</li><li>Denoise Strength: 0.4</li><li>Upscale by: Preferably x2</li><li>CFG Scale: 6 ~ 9.5</li><li>Clip Skip: 2</li></ol>Important:Positive prompt must contain: score_9,score_8_up,score_7_up,You can add "lineart, [artist tag], outline" for a more stylized result.Negative prompt can be empty or contain the default pony negatives.The promting format from Pony Diffusion V6 works well:<pre><code>score_9, score_8_up, score_7_up, just describe what you want, tag1, tag2</code></pre>If you haven't already, I highly reccomend reading through the <a target="_blank" rel="ugc" href="https://civarchive.com/models/257749/pony-diffusion-v6-xl">Pony Diffusion V6 XL model description.</a>I use <a target="_blank" rel="ugc" href="https://github.com/lllyasviel/stable-diffusion-webui-forge">Forge</a> as my main UI because it has so many optimizations for running XL/Pony models on lower end GPU's despite me having an RTX 4090, generations are alot smoother. I'm not sure how this model performs on other UI's.I haven't done much testing. If there is anything amiss, please let me know. Feedback is always appreciated :).Post results and, most importantly, have fun!

Mikoshi [Pony]

Mikoshi V1.1.safetensors

v1.1

v1.0

mikoshi-pony

模型转载自：https://civitai.com/models/543369/mikoshi-pony

Files with SHA256: 69b2cbac73365cc850bbec1d9c4d01f7564ea28a318b507069b879340902e500