LoRA trained on JoseBonn's artstyle for ponydiff: <a target="_blank" rel="ugc" href="https://civarchive.com/models/257749">https://civarchive.com/models/257749</a>Artist: <a target="_blank" rel="ugc" href="https://x.com/jose_bonn?s=20">https://x.com/jose_bonn</a> | <a target="_blank" rel="ugc" href="http://patreon.com/JoseBonn">patreon.com/JoseBonn</a>Because of my lack of experience and time, the version number may be based on changes that have made certain aspects worse. As such a previous version may perform better than a later version.I'm still learning the ins and outs of the model and LoRA training in general, please leave reproducible feedback or tips so that I may improve over time.Check the "About this version" section for any quirks or pitfalls I'm aware of, this is trained with a small dataset and so it is expected to have more bias<h3 id="heading-2001">Some Q's:</h3><ol><li>What's your workflow?<ul><li>Sampler from images is done with Restarts with segments of [3,2,0.06,0.30], which are values from the paper on Restarts</li><li>Initial resolution is typically 768 x 1152 or vice-versa and upscaled by x(5/3), I try to keep the pixel count around 1024 x 1024 for initial gens and adjust the upscale ratio as needed</li><li>Upscaler is usually 4x_fatal_Anime_500000_G or 4x_NMKD-Siax_200k</li><li>Second sample pass is the same as the first (params are on civit's meta interface) but at ~.35 denoise</li><li>A facedetailer is used</li><li>I use ComfyUI, sorry</li></ul></li><li>What strength should I be using?<ul><li>1</li></ul></li></ol>

JoseBonn - PonyDiffusionXLV6

Initial Version - Lower epochs with similar performanceMore volatile than 60E, but more flexibleStruggles with detail - jagged lineart is part of the styleComplex backgrounds lack cohesion

Files with SHA256: a2f7be66f7d3a72384e1b3d42a383d555f9ec387df7d393bc159aac0b739207d