The model I think is fairly self explanatory.
Might as well post this here since I'm sure you'll see it because the notification system is broken and you probably didn't get the alert, but on the previous model, I'd say between your two reference images there's significant difference between the two, but you'll need a third party to confirm that because maybe I'm just seeing what I want to see.
It would be curious if the trigger words did have no affect though, because I think it was two days ago I got a comment that I shouldn't use my username on all models because it skews the image, and I figure that's likely due to them mixing my lora's; and don't understand my sort of 'two trigger word system' where you can drop the malebolgia tag so that the model is more flexible and easier to mix because it's the priority word of the lora, so any 'overtraining' get's baked into the malebolgia tag and you can generally drop it without losing much of the style.
It's also a sort of experimental thing out of curiousity to eventually sort of have a way to confirm who is mixing and merging my lora's because that keyword would be baked in, because normally it would have no affect on it's own because it's not an inbuilt tag into any of this.
No I don't mind if anyone merges my stuff, and I give everyone free ability to do whatever they want with my models, It's just more of an experiment I'll test out later, because I might be crazy, but I think I saw a lot of model had my stuff in them. Though that was mostly just checkpoints. And I don't want to falsely accuse anyone, or even accuse them if I find out it's true, because it doesn't matter to me, but it did sort of throw me for a loop seeing practically preview images I've genned as the sort of cover image they'd put on their checkpoint because that's just silly. Maybe not true, but I do find it funny if I've really left that much of an impact on this site that nobody ever told me about.
https://vocaroo.com/1dCKgpymuk1M
Back to my devilish tricks.
Works well at a strength of 1
Previews made with Anima Base Model 1.0