Recommended parameters:
Weight: I myself adhere to somewhere between 0.9 and 1.1. All depending on other prompts, lora's and models.
Base trigger: numbersuit
Additional:
yellow eyes, glowing eyes, intricate iris, eyeball, eye focus - For the most iconic Number look, feel free to experiment with leaving out everything aside from 'yellow eyes', but I found what I list to work best for me.
two-tone bodysuit:0.9, blue bodysuit:0.9, light blue bodysuit:0.9 - It can be a bit fickle with coloration, on bare bones prompts I find that this isn't necessary, bit helps when going a bit further down into use of other lora's.
No specific advise on model or VAE as I've found it to work decently with most of what I've tested, I can only advise experimenting. Though I do generally stick to Clip Skip 2.
I generally also recommend taking a gaze at the added images for further prompts I have used to get the results shown.
Training data:
Trained with AnyLora on about 71 images.
