Based on my Centerfold Flux 5, trying to merge in Chroma and Krea. Diffuser version 2.3 <a target="_blank" rel="ugc" href="https://huggingface.co/Tiwaz/CenKreChro_V2_3">https://huggingface.co/Tiwaz/CenKreChro_V2_3</a>For version 2.3 I will add the quantizations under the same "tab".Big thanks to Spooknik for making the SVDQ versions <a target="_blank" rel="ugc" href="https://huggingface.co/spooknik/CenKreChro-SVDQ">https://huggingface.co/spooknik/CenKreChro-SVDQ</a> Q4 KM works fine in Comfy UI but not in Forge for example. If you need the diffuser version: <a target="_blank" rel="ugc" href="https://huggingface.co/Tiwaz/CenKreChro">https://huggingface.co/Tiwaz/CenKreChro</a> (not 100% tested yet). You can try using it as low as 12 steps even if I prefer 30. <img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/73b29f51-784e-441f-aec9-5cda12966dac/width=525/73b29f51-784e-441f-aec9-5cda12966dac.jpeg" />

CenKreChro

cenkrechro_v10FP8.safetensors

v1.0 FP8

v2.5

v2.3 SVDQ FP4

v2.3

v2.1 FP16

v2.0 Q4 K M

v2 FP8

v2 FP16

v1.0 SVDQ FP4

v1.0 NF4

v1.0 Q4 K M

test model, done for using FORGE UI, i dont test on other UI more artfull and less NSFW, if you want NSFW use special LORAs - it works fine. <a target="_blank" rel="ugc" href="https://civarchive.com/models/1295758/nsfw-fluxorwan-22orqwen-mystic-xxx?modelVersionId=2009929">https://civarchive.com/models/1295758/nsfw-fluxorwan-22orqwen-mystic-xxx?modelVersionId=2009929</a>all examples and prompts are opened in works of mine and my dear friends in AI

Files with SHA256: d2c4369287ea04e89dc4739aab9832e8c0c6964b247c7b7708af375c119565fd