toonv2(flacckly2)

used mostly xperoEnd1essmodel,but also protogen and chilloutmix

trained further with many more pictures, as a style to be used with many different models, its nsfw and regular. in settings i have tried, with both vae on and of and

*Enable quantization in K samplers for sharper and cleaner results. ON, maybe this gives better results?

You need to experiment, with this lora also,which i like