I created this LoRa because I've noticed how even if you generate images of a person in heavy rain, their faces remains completely dry - even though the rest of them are soaking wet.

This is the first LoRa I've created, and it's trained on EpicPhotogasm using Google Colab.

The LoRa seems to work best at the weights 0.4-0.7.

Without LoRa

"Portrait photo of a beautiful womans face in the rain."

With LoRa

Portrait photo of a beautiful womans face in the rain. <lora:rainface:0.6>

This LoRa was mainly created as a test, and I plan on creating a similar one for SDXL as well.