Good to see you! Try the new version!
(5/11/2005)
Today we newly released v7.0,
Sample images are generated at 720 x 1280,( Upscaler is R-ESRGAN 4x+ Anime6B 720x1280 to 1080x1920) step 25, Euler a/SGM Uniform. ADetailer and CD Tuner available, VAE is baked in. Basically not necessary,
サンプル画像は720x1280で生成されています（アップスケーラはR-ESRGAN 4x+ Anime6B 720x1280→1080x1920） step 25, Euler a/SGM Uniform. ADetailerとCD Tunerを利用、VAEは導入済み。 基本的には必要ありません。
LINK. TensorArt. yodayo. SeaArt. pixai, x.
How to use this model well
Euler a / SGM Uniform. CFG scale4-6. step20-40. / 720x1280◎. /768x1344, 896×1536/ 1080×1920 /clip skip2.
prompts
masterpiece, best quality, amazing quality, very aesthetic, absurdres, highres, newest, 2023, 2024,Additional Prompts
HDR, 8K, high detailed, subtle, high contrast, detailed background, color depth, chiaroscuro, impasto, smooth shading, glossy surfaces, shiny skin,Also recommended
dak depth, light, low kye, intricate details, masterful shading, detailed background, chromatic aberration, cinematic composition, blurry foreground, blurry background, moody atmosphere, muted colors, etc.Negative prompts
lowres, worst quality, bad quality, bad anatomy, jpeg artifacts, watermark, bad finger, missing finger,Models used for merging
coco-style-Illustrious-XLｖ1.0(master-v1.0)
coco-Illustrious-XL(v5.5,v5.6,v6.0,v6.1)
prototype model was created in the process of exploring a new style. It is still rough around the edges, but I promise it will get better.
I am good at coloring, Lighting, and artistic expression. I am also good at portraying people and SFW/NSFW.
I also aim to generate models that do not rely on LoRA or artist tags as much as they should.
And if you like it, please download it, write a review and press the Like icon. All of this will be my motivation.
To prevent unauthorized reproduction and for the benefit of those who wish to build LoRAs, we will be publishing models on other sites in the future. Reproduction of my models on websites by anyone other than myself is prohibited, even if it is not for profit. The sale of any products is prohibited. (01/21/2025)
https://huggingface.co/OnomaAIResearch/Illustrious-xl-early-release-v0
Description
Concept
This model was created for a 3D style. I tried to make the skin texture as smooth as possible, with textures closer to animation than full 3D.
It took more than three weeks to complete the project, but I am glad that I was able to release it to the public.
Features
If you don't use LoRA for the artist and character tags, I recommend a CFG scale of 4-6.
Notes.
Some artist tags may not be 3D. If only the character tag is used, the effect of 3D taste is minimal
1. How to use
1) For reforge, turn off Advanced Model Sampling for reForge.
2) CLIP is 2
3) samplers :eular A / SGM Uniform
4) CFG scale2-7. step20-40
5) I did not use various features such as upscaler, face correction, ADetailer, etc. so the detailed settings are your choice.
6) No VAE required.
7)Have fun!!
prompts
(masterpiece, best quality), amazing quality, very aesthetic, absurdres, highres, newest, safe, 3D, 3D make, HDR, 8K, high detail RAW color art, intricate details, bright face, large pupil, eyelashes, brilliant skin, shiny skin,Negative prompts
lowres, worst quality, bad quality, bad anatomy, jpeg artifacts, old, oldest, watermark,コンセプト
このモデルは3Dスタイルを目指して作りました。完全な3Dよりアニメに近いテクスチャにして、肌の質感をなるべく滑らかなになるように心掛けました。
制作を始めて３週間以上かかりましたが公開できてよかったです。
特徴
アーティストタグ、キャラクタータグをLoRAを使わない場合はCFGスケールを4-6がおススメです。CFGのスケールが低いとリアスティック寄りになります。
注意事項
アーティストタグによっては3Dにならない事もあります。キャラクタータグだけなら3Dテイストの影響は少ないです。
FAQ
Comments (53)
2D V2.0 very goooood
Awesome Merge!
this is one of the best Checkpoints i ever had
thanks for that. this style is unique
Just wow.. First time I tried Illustrious, coming from Pony, this is the next level stuff.
I want to make a lora character, what base model can I use so that the 3d effect of the 2Dv3.0 model is preserved?
great model💯👏👏👏👏👏🔥🔥🔥❤
Really impressive model
masterpiece,
This 5.5 is a masterpiece omg.
So does this have vpred from the noobai baked in? Like will https://civitai.com/models/987736/noobai-xl-detailer work on this model?
Очень жаль, что вы собрались уходить с civitai. Мне только понравилась ваша модель. Сейчас почему-то очень много авторов покидают Civitai.
Announcement.
I have discovered that v5.5 was uploaded to SeaArt without permission as soon as possible. It is a different user who is pretending to be me. I have asked SeaArt management to remove this user. Depending on the outcome, I will consider making the model private and deleting it.
アナウンス
v5.5が早速SeaArtに無断にアップロードされているのを発見しました。私の名前を騙る別人です。このユーザーを削除するようにSeaArt運営に依頼しました。その結果次第で私はモデルを非公開、削除を考えています。
For FUCK SAKE, dont disapear or delete your stuff, if yu do that, those fucker win. Dont let this down your morale at 0, you doing so great and it will be a great loss for everyone here.
Its sure annoying and disgusting feeling to see someone impersonate you but dont let this kill yur motivation and PROVE THOSE FUKER who is the real BOSS HERE.
Honestly, COCO, I can’t think of any other model besides yours. Your work gets more incredible with every new version! These past days, I’ve even been feeling a bit paranoid thinking you might delete or make it private at any moment LOL.
Don’t let these things upset you. What you’ve done with this model is a truly UNIQUE masterpiece, simply AMAZING!
Really good model only ILL non vpred I use
Announcement.
One progress: the models v5.5(-) and v5.5 reprinted in tensorart could be deleted. However, 2dv1 did not comply. The reason is that they want to do online training due to low GPU performance.
I will spare you the details of the exchange out of respect for their privacy, but it is very sad. I can sympathize with him. I too have a used desktop with a 6 year old GPU. But if we allow one person to do this, we will lose sight of the world and there will be no rules.
We will continue to respond.
Also, SeaArt was due to a bot. It seems to be collecting automatically. They are transferring this account to me.
Also, I will upload models to other places besides civitai to prevent unauthorized reproduction in the future.
I am tired......
アナウンス
1つ進捗があります。tensorartに転載されたモデルｖ5.5(-), v5.5は削除する事が出来ました。ただし、2ｄｖ1は応じてもらえませんでした。理由はGPUの性能が低いのでオンライントレーニングをしたいとの事です。
やり取りの詳細は相手のプライバシーを尊重して割愛しますが、とても悲しい。彼に同情はできます。私も中古のデスクトップで6年前のＧＰＵです。でも一人を許せば見境がなくなり、ルールがない世界になります。
引き続き対応していきます。
あと、SeaArtはボットが原因でした。自動収集しているようです。このアカウントを私に譲渡するとの事です。
あと、今後は無断転載を防ぐ為にcivitai以外にもモデルをアップロードしていきます。
疲れた.....
When I think that many of the people who are reprinting without permission are not doing it for profit, but for online training, I wonder if I should forgive them. Makes me wonder what the right thing to do is...
I've been using AI image generation models for about two years solid now. I can say without a doubt this is the cleanest model I've ever come across. Bravo.
Additional Announcements
All models that were reprinted on SeaArt have been retrieved, and as for tensorart, we are in the process of having tensorart management take care of it.
Also, the excuse was that the BOT has a partnership with civitai and SeaArt, so we can't prevent it (we're not generating any profit, so please bear with us).I didn't know. I was the one who didn't know. So it's my fault for not knowing.
Done a few Finetunes of it and some Loras. As Commissions, works good as base model.
announce
We have successfully recovered the models listed on SeaArt and TensorArt. Also, they are going to impose a penalty regarding the users who were making revenue from my models.
I wanted to tell you this last night, but I was tired and went to sleep
アナウンス
無事にSeaArtとTensorArtに掲載されたモデルを取り戻すことが出来ました。また、私のモデルで収益を上げていたユーザーに関してはペナルティが課されるそうです。
昨晩のうちに伝えたかったのですが、疲れて眠りましたｗ
In my case, when selecting the 5.5 or 5.5(-) it downloads the same file.
Fantastic! Thank you!
I never post comments on here, but this is the ABSOLUTE BEST MODEL EVER. I don't know what or how you did it, but it's literally amazing. Please don't ever take it down. It's just effing amazing.
Thank you for all of your amazing, hard work!! I would love to know how to create a similar model to fit my needs if you ever feel like helping.
Question (using V5):
I'm trying to use the furry aspect. I take for example an anthro female she-wolf and no difference - anthro horse, ram, sheep Male. And almost always I can't make the male dominate. It turns out either the futa or the she-wolf becomes a wolf ignoring the negative. The she-wolf can dominate, it does so with no problem, but it's almost impossible to get guys to dominate the she-wolf. Tips / ideas?
The colors look washed out when I use this model in comparison to WAInsfwIllustrious, why's that?
Is there a tag i need to use to increase the saturation?. Because it tends to come out washed out and the shadows looking too dark.
Any chances for an update? One of my favorite models.
I am currently working on a new model, but I am having trouble getting it past 5.6, so I am trying. I may release v.6 this weekend as a test model.
Hello! Civitai finally added "NoobAI" category. Please change on it.
It took more time than expected to adjust the quality tags and coloring, but it is finally done. If the cover image and illustrations are ready, we will publish them tomorrow morning, JST.
Thank you for the hard work! I think in general 5.5 is your best model right now!!! I am testing 6.0 version
Where are the new version files?
Bear in mind that CLIP skip 2 is not needed for any SDXL derived model, including this one and NoobAI-XL.
CLIP skip 2は、NoobAI-XLやこのモデルを含め、SDXL派生モデルには必要ありません。
the 6.0 version is illustorius or noobai it says base model illustorius ?
Fair AI Public License 1.0-SDは追加含めて条項の改変が禁止されているのでNoobライセンスとの併用はできませんよ
Illustrious派生としてアップロードするのは規約違反なのでNoobに変更することをお勧めします
Is this model trained with the COCO segmentation dataset?
very good and varied model! LoRAs work well on it too! i love it!
We will be uploading several new models this weekend. One is a continuation of the traditional taste. We have already published models outside of Civitai ( TensorArt. yodayo. SeaArt. etc...) models that have already been published on generators other than Civitai. The other will be a new taste model that combines colorful expression and delicacy by making good use of tags.
Yay! Thx for the new version!
Not really seeing it with these new versions. They just seem very deep fried?
6.2.0 and 6.0 Feel like completely different model, t's vastly different like it's not the same model at all, when using artist tag, i hope there will be an actual update of V.6.0 ,
For some reason my preferred workflow that Ive been using for Coco is not working for 6.2.1. After the efficient loader and ksampler I do a Iterative upscale but the end result have way too saturated colors and a lot of brilliant spots like particles but bigger.
Can anyone share a working workflow for this checkpoint please? Thanks!
BTW Coco from 5 to 6 has been my most used model by far. Truly a masterful model!
Love Your models man, very cool.
i loved ur 5.6 and 2.0 version i dont know how hard 2.0 was but i loved this style it was good out of sometimes adding fingers but hey it was good. i dont know if u want to build a 2.0 version in newer style but this would be cool if u can update 2.0 in this style like 5.6/6.0 cuz 6.0 was rly not that bad if u have the right setting so i add this comment as idea maybe u find a nice improvement for that . one little tip : check short words if the char will be genned or not cuz many ppl try to uses the noob ai feature in ur model to gen more characters thats my reply wish a better 2.0 in form 5.6+ if its possible stay fresh and dont give up
https://civitai.com/posts/15821637
After Tone the merge down with WAI NSFW Illus, there is the result. Amazing
Отличная модель, спасибо за обновление 6.1.
Вчерашние модели реально были не очень. С кем не бывает. Берегите себя и свое здоровье!
6.1は（また）美しいイラストを生成できます。モデルを作っていただき、ありがとうございます！
6.1 can (again) produce beautiful illustrations. Thanks for creating this model!
6.1 version good stuff<- keep this up this version is very good. protype is still alpha i think i sendet my review in the review i dont comment this here ;)
I'm using Prototype, it's insane! A Godsend, a really good and useful model for what I do!
お知らせ
今週末にIllustrious-XL-v2.0をマージしたモデルをリリースします。その関係でプロトタイプはこのページから削除され、新たに立ち上げるcocostyle-Illustrious-XLと統合される予定です。
分離する理由としてNoobAI-XLベースの6.1までとIllustrious-XLベースのプロトタイプを一緒にすることはライセンスに関しまして望ましくないからです。
画像の投稿はお控えくださいませ。そしてせっかく投稿してくださった方へごめんなさい。
プロトタイプはα版からBATA版に移行します。
α版から、ディテール、構図、ポージング、カラーリングの安定感は大きく改善できたと思います。
期待を裏切らないように今後もより良い調整モデルを提供できるように努めます。