(Please forgive the mess, using Civitai to host these LoRAs with different training settings to compare.)

V 1.081 at about .75 strength in this current batch seems to be the sweet spot. Will be dialing this in a releasing another version in not too long.

--

V2

This model continues to get use and things. I couldn't find the original dataset from back in the day so pulled something new. Still getting my training process dialed in. Will likely do a fast follow into a V3 based on what I'm seeing.

V1

Helps apply eye shadow, big lashes, etc, for that egirl vibe.

Trained on CyberRealistic Pony so it may not work well with all models.

A good prompt snippet to include might look something like this:

egirl, eyeshadow, big eyelashes, eyes, detailed eyes, close up, freckles