More realistic underwater faces.
v2.0:
Hopefully an improved version over previous one.
Recommended strength 0.7-1.0 depending on the prompt and model.
V1.0:
This is my very first attempt at creating a LoRA but it seems to work.
