Use this form to make requests: LoRA requests

V2 is less overcooked and produces imagery closer to the original style, while also being trained on 768x768 images rather than 515x512 images.

A different type of model than what I've previously uploaded based on the artist Noyr. I think they're underrated and if you agree please support their work.

This LoRA is very experimental but I feel it still works well enough now to upload it for people to try. I've used a strength of 0.7 for all of the images uploaded.