Just a friend of mine, a nice and sweet girl.

She also has a small YouTube channel.

But in general, this model is designed to diversify the resulting model generation result! since everyone sees and understands perfectly well that without using additional LoRa models, the results produced with the faces of female characters are very similar to each other.

I recommend setting the strength of the model to 0.85.

At the same time, maximum similarity is achieved.