Here is another LoRA release in celebration of my Stable Diffusion Black Book for SD 1.5 (now available on my Patreon).
This LoRA was trained on hundreds of full color, uncensored manga and doujin page images. It increases the chances of generating manga art style explicit images that have multiple panels. No grayscale images were included in the training set, and most of the training set images also had panels.
The minimum image size of the training images was 768, so it should work best around that resolution. And it was trained using the SD 1.5 base model for widest compatibility with other models.
Recommended prompts (positive):
hmanga
hmanga, uncensored
hmanga, panel
hmanga, panel, uncensored
Recommended epoch is 10 and recommend weight for most models is 0.8, but feel free to experiment. Higher epoch versions could work better on your preferred models. If images become too incoherent or 'burnt' just change to a lower epoch or weight.
Please like and follow and feel free to ask questions. And please consider following me to stay updated on my progress and releases. The Stable Diffusion Black Book for SD 1.5 is available on my Patreon along with an ever-growing collection of useful or otherwise interesting things for Civitains and other SD users.
https://www.patreon.com/slashkey
https://twitter.com/slashkeyAI
Description
Trained for 14 epochs.
FAQ
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.