Join Discord
✨ Try the Model for Free! ✨
I’m hosting my model on SocialDiff.net—a project I’m developing in my spare time. ADetailer is fully integrated, making your experience even smoother.
💖 Show Your Support! 💖
If you enjoy the model, don’t forget to leave a like and a review—your feedback helps improve the project! Also, support the talented creators publishing their work with this model by giving them some well-deserved appreciation.
🌟 Digging my way to the Illustrious Checkpoint. 🌟
Why Illustrious?
• Exceptional Attention to Detail: Every aspect is thoughtfully refined.
• Beautifully Detailed Bodies: Impressive proportions and structure.
• Perfectly Crafted Hands: No awkward poses or inconsistencies.
• Illustrious Vision for Semi-Realistic Results: A balance between realism and artistic finesse.
Key Settings
• Sampler: Euler a
• Steps: 25
Tags to Highlight Quality
masterpiece, best quality, amazing quality, very aesthetic, detailed eyes, perfect eyes, realistic eyes
Additional Enhancements
• All images are polished using ADetailer (face_yolov8s) for flawless face results.
Description
Improved body proportions
improved textures
removed extra gloss effect
FAQ
Comments (30)
Amazing work out there! Love the model and is simply wonderful. <3 Works amazing even with medium VRAM and is pretty flexible with the generation process.
Good job!
V1 - excellent results on my prompts
V2 - understands the prompt better, but my results are worse, and I can't repeat the examples,
the models from the examples and the model on civitai have different Hash ( 1620b6d479 vs f04173ac6b )
Hey, sorry for inconvenience, you need to update the model files (download it once more from Civit),
there were some errors with previous Checkpoint so i re-uploaded it.
@GZees I downloaded the model again and although the hash hasn't changed and I still can't replicate the examples, the magic is back! Now it's perfect again with my prompts! Thanks a lot!
@johnsonp33 Glad to hear that magic is back, not sure about the hash, and replicating the examples, if you could provide more details in DM.
Love both, but v2 is definitely my favorite! Quick question: I use the Draw Now app for generation, and it works fine with SDXL LoRAs in my images. Is that normal? (Sorry, I’m new to this and still figuring it all out!)
Hey, sorry not familiar with Draw Now, but if it works for you and you are happy with the results, there shouldn’t be a problem, just experiment till you find your best :)
@GZees It Does and I am happy with the results! I was wondering weather SDXL and Pony loras should even work with the checkpoint 😅
@ducktopuss964 in automatic1111 it doesn’t show non related loras for selected model(at least for me as I do remember), but you can try generating same image with same seed, one with lora and another without and see the outcome
A very beautiful Illustrious model that invites you to try it out. Well done
This makes the nicest Illustrious images on the site; how can we get this hosted for On-Site Generation ?
I don’t know how it works, all depends on the popularity of the checkpoint I guess, so more likes and downloads I think
@GZees let me know if I can help; my personal computer does a bad job with Illustrious, so o need to do in site gens!
@pornomyc i'm currently working on a platform where you could use my models, same setup i do. also character builder, so stay in touch :)
How is the facial feature variety? A lot of the example images and in the Gallery are all very similar.
Facial variety is alright, it's adetailer that makes all the faces similar, also the prompt it's not changed a lot from my side
@GZees I've had slightly better results using face_yolo8m but I don't know if I'm imagining it.
@turkino a large number of the faces look the same in this gallery in part because many people are using the same loras ( <lora:MoriiMee_Gothic_Niji_Style_Illustrious_r1:>) in particular. I also think there's a secret art to using illustrious models that few people really understand (myself included), that uses artist referencing in the prompts. Illustrious models use a "booru" based data set for the most part. It's why they're so effective following prompts, but struggle with things outside of anime. But this booru base set will have plenty of artist tags in it. So if you know your doujin artists, you can try and prompt some of the characteristics of their work out of your results.
It can help to throw a random name into your prompt, like "amy", "alexis", "emily", "damien", "george", etc when trying to create a distinct, consistent character. It's hard to get TOO far away from same-face with anime/cartoon characters but this helps a lot. You can also do something like (amy:1.5) in your A-Detailer prompt to not lose the distinct features of the character. Play around! Mixing that with the tip to use artist tags can give you great control over characters.
After a lot of "illustrious is the new pony" hype, its nice to actually see a model that can do western art styles effectively. Still experimenting with it but so far so good!
Model is really good.
Testing some prompts for diff and original characters... posted.
Thumbs up mate.
Sure it is the best model I tried so far.
Overwhelming quality.
This is one of the only Ilustrious base models I have had any success with!
This one needs to be available for on site generation instead of those ones with flat styles.
Hey, till then you can use https://socialdiff.net
Hi! Can you please tell me if I need a VAE? If so, would you be so kind as to tell me which VAE you are using? And if it's not too much trouble please give me a link. Thank you very much in advance for your reply.
Amazing model, and huge props for hosting it for free on your website! So much fun and creativity packed in there!
Model is a hidden gem, knows all the characters from NoobAI but generates in a semirealistic style.
I recommend the creator adds some tags so people can find it and they have more exposure.
Where is it possible to find info about the positive and negative prompts: <lora:MoriiMee_Gothic_Niji_Style_Illustrious_r1:0.5> artist:moriimee, <lora:PowerPuffMixLora:0.4>,
these are LoRAs (just search their names on this website): https://civitai.com/models/915918?modelVersionId=1244133
I love this checkpoint, low key obsessed :D
Details
Files
Available On (3 platforms)
Same model published on other platforms. May have additional downloads or version variants.