The Agenda Mix
Designed by Ktiseos Nyx.
We've been at this since January of 2023! If you're interested in supporting us in our long standing AI journey please scroll down for ways to connect & sponsor us! Currently if you're interested in having it on the generator please feel free to donate towards auction costs!
✨ Recommended Settings
To get the best out of ANY of our models, try these settings:
Samplers: Euler or Euler Ancestral
Scheduler: Karras
Negative Prompt: Generally not needed! The model is designed to perform well without one, but feel free to experiment based on your preferences.
HiRes Fix:
Steps: 15
Denoise: 0.3
Upscale: 2x
Upscalers: Rex, 4x_Remacri, Lollypop or your personal favorite.
ADetailer: Use sparingly. Due to potential VAE differences, it might occasionally produce grey squares.
Embeddings: We've used some of our own custom embeddings to showcase character types in our examples, but no stylistic embeddings were necessary to achieve the core look.
Checkpoint Compatibility: Depends, not everything will behave.
🛠️ Training Details
If you're curious about what's the training details in full make sure to use this tool: https://xypher7.github.io/lora-metadata-viewer/
🙏 Acknowledgements
A huge thank you to the incredible community members whose work and inspiration contribute so much. Special mentions to:
Arthemy, Wai, Holostrawberry, FallenIncursio, RichyRich, JustTNP, Hoseki, CitronLegacy, Samael1976, Kairen92, ShadowxArt, Flyx3, YeiYeiart, LazyVisionary, and so many more! We try to train as much as we can ourselves, but the community's creativity is a constant source of inspiration.
🌟 About Ktiseos Nyx & Connect With Us!
Ktiseos Nyx is available for commissions! Find us and more of our creations here:
Discord: https://discord.gg/HhBSvM9gBY
Our Other Socials: https://beacons.ai/duskfallcrew
Sponsor Models & Commissions: https://ko-fi.com/duskfallcrew/commissions
TensorArt Space: https://tensor.art/u/611011406535381539
Dataset Tools: https://github.com/Ktiseos-Nyx/Dataset-Tools
💵 Training Costs Money!
While Civitai style trains usually cost buzz, testing and infrastructure to create better quality content to service our customers long term does cost money.
We're not looking to be the next Civitai at all, but AI has been a format of income outside of the inability to find regular work.
Donations, Commissions and Sponsorships are the backbone of our work!
Buzz donations work to keep our models on the generator that YOU the user choose! (This was originally a rarity, buy recently you've all chosen and we're all proud!)
Lora training can take time and effort, and while we do have cost layouts - if you're wanting a simple character lora you can use the Ko-Fi link shown in our link In Bio Above!
✨ Commission Details
If you're wondering why you're seeing a style lora or other loras ahead of your queue'd training or commission please see these FAQ lines:
Style Loras are usually developed independent of our normal training queues at the moment, as we use Civitai's platform to train them for use with our normal model making schedule during the month.
Character loras normally are trained as they come in, these semi depend on our own training service we've developed. Right now the NextJS system is almost complete for testing, and we will be resuming PAID and Free queues ASAP.
Concepts that are CLOSE to "STYLES" are again created using Civitai, at least on the short term. Once our main service has been completed we will amp up our schedule.
SAMPLE IMAGE AND PROMPT:

(masterpiece, best illustration, extreme light and shadow), 1boy, solo, (male focus:.7), transgender male, red hair, white hair, green eyes, two tone hair, jacket, green jacket, gloves, green eyes, white hair, headband, bodysuit, yellow gloves, yellow bodysuit, multicolored hair, black headband, looking up, multicolored clothes, hair behind ear, long hair, green bodysuit, looking at viewer, open jacket, yellow jacket, arm up, multicolored bodysuit, goth eye liner, wristband, emo punk style, (punk:1.5), jewelry, depth of field, tattoos, minneapolis, (superhero:.4), (rogue of the x-men:1.2), wind swept hair, (seductive), (sidelighting), (volumetric lighting)
best quality, masterpiece, ultra high res, photorealistic, 1girl, offshoulder, smile (that's it that's the prompt, someone on the bug fixes site wanted a gen lmao - this was their prompt XD)
(masterpiece, best_quality, ultra-detailed, immaculate:1.3), epic, illustration, render, volumetric lighting, welcoming, Finger Lime entity, see-through gossamer, , in Japan, striped elbow gloves, neon green color scheme, multicolored hair, bombshell hair, rainbow hair, half updo,Blowout Curls, upside-down<lyco:EnvyRadioactive01-sd15-d:1> <lora:Niji_Slime:0.3>
long hair, looking at viewer, 1boy, brown eyes, jewelry, blue hair, collarbone, flower, male focus, earrings, solo focus, water, ring,Description
FAQ
Comments (44)
Awesome! Thanks for making this. Glad to see some trans positivity here :)
I thought of you when I was doing the prompting, I know general chat gets a bit weird in the discord- and I ALMOST PROMPTED PROGRAMMER SOCKS CAUSE OF YOU! :P - You're welcome in the Earth & Dusk discord any day <3
This checkpoint is super loud and bombastic, which I vastly prefer to "incremental waifu merge that's 0.5% different", so good job.
Question: Was this actually trained on trans characters, or are those tags in the prompt largely ceremonial at this point? Either way's fine, I understand booru tags aren't exactly "inclusive".
No, sadly but i DO have a Lycoris that was trained on a Transgender female that is an alter in a friend's system - It's based off their second life avatar, and at a varied strength would work well with this model, would you like a link?
As for "CEREMONIAL" - Like, it's not I THINK IN MY EYES cause it's an anime style model, you can't really get PERFECTION in how you represent gender diversity, and no i know you're entirely being kind and querying - so fear not i'm not discrediting you at all - The prompts wre FULLLY INTENTIONAL as well as possible, so I GUESS yea its accidentally ceremonial - HOWEVER one of the backwards models in it way back DOES Now that i think about it, in the Duskfall Alters model set - a trained one of ours "EPIC TRAIN" is what started these merges, so all incidentally of the ones that have "EPIC MIX" of any sort are IN THEORY Backwards compatible LOL, it doesn't mean the mixes have kept true to that, but i promise you the intention is there :3
So for example this is the mix that started this all over like a week or two ago: https://civitai.com/models/20562/epic-mix-anime-realism-and-nsfw-support (We have DID/CPTSD/Autism and ADHD - if this was a week ago then i'm insane XD) - It has "EPIC TRAIN " in it which was a fine tune model using Epic Diffusion, and a ton of self made resources - I also have links to that as well as the lycoris :#
Lycoris with trans people in it: https://civitai.com/models/18163/visual-3d-lycoris-lora
Other one: (https://civitai.com/models/16763/virtual-worlds-lora-set Lora) https://civitai.com/models/25430/virtual-worlds-people-lycoris-set
Epic Train with trans people in it: https://civitai.com/models/11565/duskfall-alters-portraits
What VAE do you recommend with this?
I use Kenshi, but some people may prefer counterfeit, blessed or other?
safetensor? Please...?!
SOOOOON <3 I'll. be getting that done today maaaybe but YES YES - Safe tensors soon <3
Ok it's up at HF: https://huggingface.co/Duskfallcrew/AgendaMix but i'm downloading it NOW to upload to civit :3
I love these color schemes. Better pastels than pastelmix!
Well Pastel mix is largely a STYLE, where as this was a weird mix that wasn't originally intended one way or another - and alot of it's not pastel per se it's - just what it is XD I mean it's a merge, and Pastel mix is a Anything type merge, and i think Anything 4 or 3 is in the mix somehwere in this XD it's --- you'd have to see the bucket mix for the derailed details XD I can never remember XD
Wow! just perfect. Would it be possible for you to create a Lora/Lycoris model of it?
I don't know how to extract them no, BUT I might try and take the sample data i provided and make a style lora with it :3
@duskfallcrew will be easy on the harddrive :) but loving the style. Very beautiful. Keep up the great work mate.
@okurin2018242 Sadly, LOL lora extraction on my hardware is NOT possible :P however, i wouldn't mind trying to make a style lora out of the samples i've curated and made :3 - other than that can always beg the extraction gods :#
I think I understand now how to extract one, as soon as I figure it out i'll do one of both the new and old versions <3
EYYY SO Uh we didn't extract one from 1.0, but we DID FINALLY TRAIN one on 1,0's outputs!!
Out of my four amazing kids, one is a Trans Man with D.I.D., and one is an Enby. I'm a middle-aged Trans Woman, and proud of it, and incredibly proud of them too! 😍
Thanks for promoting inclusivity, as it should be promoted everywhere...
As always, and some would say it's "SAD" they have DID. - and in some ways it is - because the trauma endured to create it - is not one anyone would want to fare with - And that aside, i'm aware other plural types exist, but merely i'm just speaking from our experience smiles - We've been told to take up more space, and as one of the Adult aged offspring of the main fronters, I promise you we'll continue flying flags on Civit <3 especially since the mods and staff allow it and continually tell us to report comments that are disturbing, homophobic etc <3
@duskfallcrew He has good days, and ones where he's struggling... I'm working on getting all four away from their current toxic environment. This should happen in a year, fingers crossed x
anyway you could trim down the model? its a little fat
Yes 👍 I can prune it soon, was struggling to download it from vast somi focused on the larger one 😍
@duskfallcrew sounds good
@duskfallcrew how much was DreamShaper added into this?
@LostAtSea none lol, 🤣 absolutely none
@duskfallcrew update?
@LostAtSea FLARK forgot LOL - XD I'll get this done soonish when i get some other stuff done - Been forgetting a lot <3
FINALLY GOT YOU A PRUNED EDITION, ALWAYS poke us about this- We get it, and we're slow at it - IT IS struggly wuggly even with OUR resources to use higher size models (God have you seen SDXL?) - Earth and I have new stuff that might be lacking pruned editions, do me a favor on ANY of our content - ALWAYS POKE ok?
You gucci.
You valid.
We gotchu <3
SOMEHOW 1.0 wasn't pruned, dont know why LOL but 1.0 is pruned and 3.5 is coming to update that it SHOULD In theory convert to diffusers now. XD
Time for an SDXL of this dope ass CP, per favore.
LOL Envy just did one that's ROUGHLY based on this, i'm REALLY BAD at msking SDXL stuff, what I CAN DO is run my own version of the Agenda Mix thing for SDXL, OR -- IF ANYTHING, I Grant you full permission to run gens and make your own SDXL lora out of any of my 1.5 checkpoints you love. I can get loras done, but i'm struggling cause i don't have a good computer- i rent GPU's do do checkpoints and so far i've just... X__X i'm not feeling up to fighting SD to get SDXL to work lol
@duskfallcrew I get it. I'm spoiled. I actually found this one due to the Envy remix, which seems a bit watered-down from this. Re: renting GPUs, I don't do Patreon, but I do Ko-fi and direct PayPal, would be happy to hit you with a few bucks for the LORAs you recently ported to SDXL.
@ZootSuit OH shweet, I -- Yea we're never closing our Ko-Fi, just that we're combining a lot of our activity into ONE for Earth & Dusk - Ko-Fi gives less fees than direct paypal, and then i don't ahve to give you my direct email XD
@duskfallcrew okay but I don't see your Ko-fi link anywhere easily discovered.
@ZootSuit OH! I must've replaced it XD Oof hang tight lemme HERE YA GO: https://ko-fi.com/DUSKFALLcrew - do check out bestie's - they dont have stuff ON THEIR KO-FI but they'll port some of their stuff ASAP to SDXL soon https://ko-fi.com/earthnicity @earthnicity https://civitai.com/user/earthnicity/models
Just updating you: Now that 3.0 is out, and I have a clue what i'm doing with SDXL loras, i'll gen some more 1, and 2.0 content and mix it with 3.0 outputs and try and use Envy's model with one of my anime models as a backmix soon. NO promises yet, i've been renting 4090s so the only thing i gotta do is TRAIN the lora before i go mixing it ;)
@ZootSuit With 3.5 out, and more than enough data to make a strong SDXL lora to bake in -- we're gonna work before xmas on agetting Agenda XL out <3
I cannot wait to try this one!
Remove the photorealistic tag
Will do, was gonna whine you had commented twice in which case i saw it on two models - I'm sorry XD - Nah these were old AF models to start with, and it was back before you COULD block tags. Thank you for your patience i'll fix that asap!
All done! Please do let me know. either in comments or DM's if any other model has this issue, I do understand it's a "I need to block anime but this is coming out in my feed" - And again it's not your fault at all, this was just a model long before we could block based on subject <3






