Sign In

Virtual Diffusion XL

52
567
19
Early Access
Verified:
SafeTensor
Type
Checkpoint Merge
Stats
224
Reviews
Published
Oct 28, 2023
Base Model
SDXL 1.0
Usage Tips
Clip Skip: 2
Hash
AutoV2
28E315FFE8

As of 1.5 - this is now Virtual Diffusion XL, Slimy mix wasn't really WORTH continuing as an update since we're TRYING TO LIMIT THE NUMBER OF EXTRA LINES we make for SDXL LOL.

This new version is largely based on the 3D/ Second Life style. We used this largely to help influence: https://civitai.com/models/275116/virtual-diffusion-second-life-xl

A lot of our SDXL trains with lora weaving use animagine, so we've included the liscence below:

We're looking for more content creators:

https://www.end-media.org

Our Discord:https://discord.gg/5t2kYxt7An

Backups: https://huggingface.co/EarthnDusk
Send
a Pizza: https://ko-fi.com/duskfallcrew/

"WE"? - We have Dissociative identity disorder, ADHD, Autism and CPTSD - "WE" as in we're a system of over 200 alters, and we're not ashamed about it. We believe that AI can break down barriers in some aspects of mental health, but we also believe that AI can hinder aspects of it.

License

Since we used Animagine XL and such alot we're literally just using this from now on: Animagine XL 3.0 now uses the Fair AI Public License 1.0-SD, compatible with Stable Diffusion models. Key points:

  1. Modification Sharing: If you modify Animagine XL 3.0, you must share both your changes and the original license.

  2. Source Code Accessibility: If your modified version is network-accessible, provide a way (like a download link) for others to get the source code. This applies to derived models too.

  3. Distribution Terms: Any distribution must be under this license or another with similar rules.

  4. Compliance: Non-compliance must be fixed within 30 days to avoid license termination, emphasizing transparency and adherence to open-source values.

The choice of this license aims to keep Animagine XL 3.0 open and modifiable, aligning with open source community spirit. It protects contributors and users, encouraging a collaborative, ethical open-source community. This ensures the model not only benefits from communal input but also respects open-source development freedoms.

WE ARE PROUDLY SPONSORED BY:

https://www.piratediffusion.com/

https://yodayo.com/

JOIN OUR DA GROUP: https://www.deviantart.com/diffusionai

JOIN OUR SUBREDDIT: https://www.reddit.com/r/earthndusk/

DISCLAIMER:

We aren't 100% sure what other models we used to stabilize this, but we do know the lisc restrictions were open on all of them.

We do know for a fact somewhere in the backmix we used SocalAniMerge - why? Because it was part with AgendaXL!

We TRIED flustercluck with it, but it somehow messed up the proportions -- which isn't Chief's fault, just SDXL is REALLY PEDANTIC on how things mesh and weave. It's not easy compared to SD 1.5

Future versions have Ratatoskr and Faeia's SDXL model, as well as future updates coming will have DPO.

LORAS that weren't ours that i KNOW we used:
Someone made an Absolute AMAZING LUT Lora, and applied to a semi real model DOES make it mostly realistic, which is TOTALLY FINE, it's amazing.

This time around apart from that and an anime detailer lora from Caligstro/Animagine XL - we didn't use a lot of non E&D loras.

IN theory i beleive itwas just the following:
MJ Osea XL (which is coming I have the pics I'm just tired lemme alone)
the SL/Virtual Diffusion LORA for XL

League XL from us
V Diff update for SDXL v2

What DIDNT GET USED:
PorcelainEarthXL
Heartsteel XL

Because weirdly their 3d/2.5D styles really weirded out on the model mixes I was using to boost this.

Also secondary disclaimer:
While in theory this is a "MERGE" - when you spend as much time and effort making a lora FOR TEH REASON TO MAKE A CHECKPOINT,in theory it's a train. it's just not traditionally trained.

I will forever balance this by calling Agenda XL a merge from now on because it just is? XD