<h1 id="overview:">Overview: </h1>First merge, please understand and have patience.This is mix between Meichi IL Light and Rimix Omega v2.All previews are bare model generations, absolutely NO postprocessing was done on any of themMy personal settings: CFG 4, DPM++ 2M/Karras, 25 steps But feel free to use what you are comfortable/happy with. Sampler choice is personal, but I recommend keeping CFG reasonable. Credits:<a target="_blank" rel="ugc" href="https://civarchive.com/models/1193368/meichi-il-ightmixv1">Meichi Light</a> by @JuzuArupukato <a target="_blank" rel="ugc" href="https://civarchive.com/models/996495/ri-mix-illustrious-anima?modelVersionId=2811854">Rimix Omega v2</a> by @phinjo <h3 id="comparison:">Comparison:</h3><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/25e872b9-892b-4a45-8a2c-072038fd8596/original=true/25e872b9-892b-4a45-8a2c-072038fd8596.jpeg" /><h3 id="versions:">Versions:</h3>V1: Initial mix, with the model weighted towards Meichi Light and the CLIP slightly weighted towards Rimix.Not all model blocks merges were weighted evenly, the inner blocks are progressively weighted more heavily towards Meichi-Light. Details on the merge can be found in <a target="_blank" rel="ugc" href="https://civarchive.com/articles/37796/experiment-failed-successfully-hypothesis-inverted">this article</a>. V2: Reverse mix with the same weighting; more model weight on Rimix, less CLIP weight on Meichi. Same merge strategy, just with the model priority opposed. THIS IS A VARIATION, NOT AN IMPROVEMENT.

Cyberpaint DS

V1, first ever attempt at creating a merge

Files with SHA256: 498eed78a6cab998fc0432a2223c7a22aa0254515dcf2927c43ec6198e04578b