>>11332549
>any activation words for the foreskin lora
There actually isn't a trigger word trained at all, most of the time the tag 'foreskin' is enough. Excessive foreskin as a tag was not tagged in the dataset but sometimes it helps to use it, feels like a little like placebo though. I like to use 'foreskin' with sometimes different weight. Also inpainting with the lora seems like the best way to get more detail on the foreskin. With picrel I inpainted the genitals because they were just too undefined in the base gen.
>brackets closing this part and the pink leotard followed by a colon
That's just the syntax for the prompt sceduling that my setup uses. I'm not sure if your interface supports it but it's a pretty neat way to influence and change tags during the genning process. You can google 'prompt sceduling' or 'Keyword blending' for more info.
>Here's the the tl;dr as a quote from a random guide:
schedule prompt terms to only apply during particular steps of the generation:
[man:12] will apply the term beginning at the 12th step of the generation.
[man::12] will apply the terrm until the 12th step.
[man:woman:12] will switch from man to woman at the 12th step of the generation
[man:woman::12]:5] will begin applying man on step 5, and switch to woman on step 12.
>[huge ass:0.1]
This prompt means that 'huge ass' gets added to the prompt after 10% of the steps while genning.
>the "full frontal" tag is "straight-on"
Lmao, no wonder it never seemed to work to neg that. Thanks for the tip.
>>11331632
>homosimile was being quite stubborn
Yeah that's why I mentioned that the specific shitmix model barely matters. Just use what works best for you hahaha I tend to switch up models too.
>>11331808
lol nice