What is this LoRa?
This LoRa has been trained on Manga covers for Japanese Visual Novels on the SD3.5 Large base model
Recommended LoRa strength: 0.5
CLIP skip: 1
The keyword for this Lora is "mangacover kei" . The prompts are created using JoyCaption:
Sampler used is DPM++_2M using SGM_uniform scheduler.
Training images+prompts:
You can download the training data for this LoRa here: