check point - 0.5(chilloutmix_NiPrunedFp32Fix) + 0.5(pornvision_final)
LoRA weight - .6 or .7 ( experiment it )
CFG - 3.5
( test out a mix of CFG vs Weight to find optimal mix for your prompts)
Note: i did a cheeky training with this model - so this will give very NFSW results even when you dont try :) -- so i suggest using 'NFSW, in negative prompt.
Description
trained on different images...