Merge model that aims for compatibility with Pony LoRAs on SDXL models.Use both SDXL and Pony prompt tags. Images are generated without hi-res fix / upscaling.Models tagged with "base" are generally worse for inference if you are not using additional LoRAs.Base Model MergesAni3130b - (0.5 x Ani30Base) + (0.5 x Ani31)GBbp - (0.5 x GammaXL) + (0.5 x BadbroPastel)JRU21 - (0.5 x JRU_v2) + (0.5 x JRU_v1)KEr3BAxl3 - (0.5 x KohakuEpRev3 [get rev3 from HF]) + (0.5 x BAxlv3)LoRAsExh - ExpressiveH Pony LoRA by jspp3733410Bs - BlueStyle SDXL LoRA by HCSJRrd - Rurudo Pony LoRA by reweikIti - Ichiri Pony LoRA by zombocromboModelsMod4 - (0.5 x Mod3) + (0.5 x KEr3BAxl3)[Mod5 final base] Mod5 test2 - (0.4 x Mod4) + (0.6 x JRU21)[Mod5 LoRA merge] - Mod5 [final base] + Exh str 1 + BS str1Mod6 MODEL - (0.5 x Mod4) + (0.5 x Bbp + Exh str1 + Bs str1)[Mod6 final base] Mod6 CLIP - (0.5 x Bbp + Exh str1) + (0.5 x Bbp + Exh str1 + Bs str1)[Mod6_rrd] - Mod6 [final base] + Rrd reweik str 0.33[Mod6_iti] - Mod6 [final base] + Iti zombocrombo str 0.40Mod7 MODEL - (0.25 x Mod6) + (0.75 x JRU21)[Mod7 final base] Mod7 CLIP - (0.5 x Mod6) + (0.5 x JRU21)[Mod7 LoRA merge] - Mod7 CLIP + Exh str 1 + BS str1ScrappedMod1 [Scrapped] - (0.5 x KEr3BAxl3) + (0.5 x Ani3130b)Mod2 [Scrapped] - (0.5 x Mod1) + (0.5 x GBbp)Mod3 [Scrapped] - (0.5 x Mod2) + (0.5 x Bbp + Exh str1)
