By gerogero
Updated: November 8, 2024
LoRAs may be the reason that you’re not getting the results you want with your NSFW generations.
LoRAs smaller models you use in addition to existing checkpoint models. They’re trained on a specific subject (such as an art style, a character, a body type, a sex act etc), and will push the generations towards that subject.
Some LoRAs only work well with a specific model (eg. chilloutmix), others work well with most models.
I would roughly divide them into the following categories:
IMPORTANT NOTE: When using LoRA models, users should abide by local laws and regulations and shall not violate any other person’s right to privacy
You use them by downloading the LoRA file, placing them in the lora folder in your stable-diffusion-webui folder, and using a trigger word for the LoRA in the prompt.
You can use as many LoRAs in the same prompt as you want.
guide to SD LoRAs, and training your own
Here are some good LoRAs that enhance the results of your NSFW models:
OG LoRA, certainly one of the most famous. Trained on Korean models, not any specific person. Works well with Chilloutmix but works with other models.
Trained on Chinese Tiktok influencers, not any specific real person. Works well with Chilloutmix but works with other models.
This is a LORA for better nude body types out of generated images: makes proportions sexier.
The idea is to make the vulva on the inside for better aesthetics.
Doll-like face LoRA
Ahegao (アヘ顔) is a term in Japanese porn for a facial expression of characters during sex, typically with rolling or crossed eyes and tongue out.
This is a beginner’s guide to help you install Wan and implement every available optimization to maximize the speed of video generation. Now ...
Niji Mode AKA Niji Journey refers to Midjourney’s anime-specific models. They are built on entirely different architectures from the standard...
The Power of Words The words we use in prompting are called tokens. Each token has its own power, which depends on its frequency of occurrence in t...
LoRAs (Low-Rank Adaptations) are smaller files (anywhere from 1MB ~ 200MB) that you combine with an existing Stable Diffusion checkpoint models to ...