This is a bit of a experiment as many of the TI's I tried failed and cost several days of work. In the end I did a bit of a mix between Subject and Style, combined with prompt mixing.
Samples generated on Protogen v2.2
How to use:
It's best to use this as a mix-in with a subject. For example, with a woman you would do (woman mixed with (tloustyle))
The model has a bias with trying to produce girls, which is odd considering there was like 2 female infected subjects trained on. Anyways, to defeat this, when doing a man, you may need to add negative prompt such as ((female, woman, girl))