Illustrious info : Innate style of the model feels not enough trained, lora should reinforce it. Probably a little overbaked though. Config with params attached for this version.

Pony info : Two versions, one trained from base pony and have some nice aesthetics, while suffering at style sometimes, second is trained from autismmix and forcing a style replication really hard, but almost corrupts small details and ???????, either use pony version with higher than 1 weight if style doesn't replicated enough or just a autismmix version for a harder style replication. Personally I like more the pony version with rare need to raise the weight. Vectorscope is also a good addition