/pkg/ AI models

Work In Progress

Model Principles

I'm not responsible for what you create with these AI models
Please read the Restrictions in the FAQ section. These are grandfathered from Stable Diffusion's Licensing. TL;DR don't be an ass!
Models will have Soft, Medium & Hard variations from different epochs. Soft ones are from earlier epochs and will be easier to change the character, even making it unrecognizable, hard ones might be over-trained but will require fewer words to just get the character on the screen.
Recommended strength weights are usually between 0.4-0.8.
My models are trained on Clip Skip: 2 so you should probably use it for best results. (but clip skip 1 might also work nice)

New Google Colab Version

if you have no idea how to use this. Just follow the stars. I still recommend the local version if you're able to run it.
https://colab.research.google.com/drive/1KmOk-ApO3wj3tidH3_Vs8NTFrwjeB3HG?usp=share_link

Guide

How to use the models locally?

1. Follow Voldy guide https://rentry.co/voldy to get Stable Diffusion UI and install it on your machine. A med to high end GPU is required(Think RTX1660/RX580 or above)
2. Update to latest version
3. After you get the UI working and test a bit with the base stable diffusion model, download additional base models and place them in <sd-ui-folder>/models/Stable-diffusion. Some examples are: Anything-v4.5-pruned, Hash:6e430eb514 && AbyssOrangeMix3, Hash: d124fc18f0. I currently use MeinaUnreal, Hash: 5a1c2b4a08
4. Get familiar with it!!!

5. Follow one of the methods below to use LoRAs(the models I added on these pages):

   Method 1 (Recommended)

   Turn on your Stable Diffusion UI.
   Go to the 'Extensions' tab.
   Go to the 'Install from URL' subtab.
   Add in the URL: https://github.com/kohya-ss/sd-webui-additional-networks
   Press Install
   Go back to the 'Installed' subtab
   Press 'Apply and Restart UI'
   You should now have a dropdown called 'Additional Networks' in your txt2img generation tab
   Unzip .safetensor models in 'sd folder>/extensions/sd-webui-additional-networks/models/lora/' (create the lora folder if needed)
   Use the "Additional Networks" tab under the image settings in txt2img tab
   Check the Enable button, then press Refresh Models, Load up one of the models in Model 1, set its weight, then prompt for it and generate an Image. You can use the prompts from each of the individual pages.

   

Method 2
Download the .safetensor models from one of the girl's pages.
Unzip/place .safetensor models in sd-webui-folder/models/Lora
Call them in your prompt like <lora:lora_name:strength>. Eg. <lora:Aletta:0.5>
Use the Additional Networks button under Generate, then click on the Lora Tab. Click on the Lora you want to use, then edit its weight in the prompt
Use trigger words in prompt to better guide the model towards the character, the names should be the same as the file name, but catbox changes it to random string ,so it depends on how you name the safetensor files, unless you really like calling them <lora:uwv6zus:0.5>

How to Use Hi Rez Fix (aka how to make even better images)

Coming soon, but in short, just set it up EXACTLY like this(with the same resolutions, don't increase them unless you have a 24GB vram GPU):

Models

All LoRA models can now be found on HuggingFace: https://huggingface.co/346368616E/Kawaii_LoRA

Lua Asuka - https://rentry.org/lua_asuka_AI

Namiji Freesia - https://rentry.org/namiji_freesia_AI

Shee Icho - https://rentry.org/shee_icho_AI

Charlotte Suzu - https://rentry.org/charlotte_suzu_AI

Isla Coleman - https://rentry.org/isla_coleman_ai

Nene Amano - https://rentry.org/nene_amano_AI

Reina Sun - https://rentry.org/reina_sun_AI

Archive

Archive - https://mega.nz/folder/jiZBEYyA#0i8LUUWNtkfCP_TyYIDO0Q

Friends

Menace Alchemilla - https://rentry.org/menace_AI

Liora Walkyria - https://rentry.org/liora_walkyria_AI

Sara Nagare - https://rentry.org/sara_nagare_ai

Graduated

Aletta Sky + Lettiverse - https://rentry.org/aletta_sky_AI

Sava Safari - https://rentry.org/sava_safari_AI

Peony Aeria - https://rentry.org/peony_aeria_AI

Miryu Kotofuji - https://rentry.org/miryu_kotofuji_AI

Additional Extensions and Reading

These can be installed from the Extensions tab in your UI
ControlNet: https://github.com/Mikubill/sd-webui-controlnet
ControlNet models: https://huggingface.co/webui/ControlNet-modules-safetensors/tree/main

Place the ControlNet models in <SD-webui>/extensions/sd-webui-controlnet/models
Restart your SD Webui

Openpose editor: https://github.com/fkunn1326/openpose-editor
Online openpose editor: https://zhuyu1997.github.io/open-pose-editor/
Pose---Click on Generate Skeleton/Depth/Normal/Canny Map---Click on the first image on the bottom right, the one with a stickman---Save it---Load it up in the ControlNet tab---Preprocessor None---Model Openpose

FAQ

Why am I doing this?
I want to see AI art accepted in a separate AI Art tag and share my own there.
Feel free to outright steal samples from here or the threads and share them on discord, twitter, call them your own, idgaf. It will probably tick you off as having been /here/

Tripcode check?
/pkg/-AIBot !!bPKjjk1ioWN

License?
I would Say this falls under creativeml-openrail-m, same license used by Stable Diffusion.
Read more about it here: https://huggingface.co/spaces/CompVis/stable-diffusion-license
In short, don't be an Ass!
In longer format: follow these guidelines:
The model should not be used to intentionally create or disseminate images that create hostile or alienating environments for people. This includes generating images that people would foreseeably find disturbing, distressing, or offensive; or content that propagates historical or current stereotypes.
Out-of-Scope Use

The model was not trained to be factual or true representations of people or events, and therefore using the model to generate such content is out-of-scope for the abilities of this model.
Misuse and Malicious Use

Using the model to generate content that is cruel to individuals is a misuse of this model. This includes, but is not limited to:

The model should not be used to intentionally create or disseminate images that create hostile or alienating environments for people. This includes generating images that people would foreseeably find disturbing, distressing, or offensive; or content that propagates historical or current stereotypes.
Out-of-Scope Use

The model was not trained to be factual or true representations of people or events, and therefore using the model to generate such content is out-of-scope for the abilities of this model.
Misuse and Malicious Use

Using the model to generate content that is cruel to individuals is a misuse of this model. This includes, but is not limited to:

Use Restrictions

You agree not to use the Model or Derivatives of the Model:

• In any way that violates any applicable national, federal, state, local
  or international law or regulation;
• For the purpose of exploiting, harming or attempting to exploit or harm
  minors in any way;
• To generate or disseminate verifiably false information and/or content
  with the purpose of harming others;
• To generate or disseminate personal identifiable information that can
  be used to harm an individual;
• To defame, disparage or otherwise harass others;
• For fully automated decision making that adversely impacts an
  individual’s legal rights or otherwise creates or modifies a binding,
  enforceable obligation;
• For any use intended to or which has the effect of discriminating
  against or harming individuals or groups based on online or offline
  social behavior or known or predicted personal or personality
  characteristics;
• To exploit any of the vulnerabilities of a specific group of persons
  based on their age, social, physical or mental characteristics, in order
  to materially distort the behavior of a person pertaining to that group
  in a manner that causes or is likely to cause that person or another
  person physical or psychological harm;
• For any use intended to or which has the effect of discriminating
  against individuals or groups based on legally protected characteristics
  or categories;
• To provide medical advice and medical results interpretation;
• To generate or disseminate information for the purpose to be used for
  administration of justice, law enforcement, immigration or asylum
  processes, such as predicting an individual will commit fraud/crime
  commitment (e.g. by text profiling, drawing causal relationships between
  assertions made in documents, indiscriminate and arbitrarily-targeted
  use).

How do I train my own models?
https://rentry.org/lora_train