For business inquires, commercial licensing, consultation and large scale image captioning for datasets contact me under [email protected], for custom LoRAs visit my ko-fi page.
๐๐ฑ๐ฎ๐ฌ๐ด๐น๐ธ๐ฒ๐ท๐ฝ ๐ฃ๐๐น๐ฎ: ๐๐ท๐ฒ๐ถ๐ฎ/๐๐ธ๐ท๐ฌ๐ฎ๐น๐ฝ/๐๐ฎ๐ท๐ฎ๐ป๐ช๐ต ๐ค๐ผ๐ฎ/๐๐ช๐ท๐ฝ๐ช๐ผ๐ /๐๐ธ๐ถ๐ฒ๐ฌ
Follow me on Twitter: @YamerOfficial Discord: yamer_ai
If you have enjoyed this checkpoint:
Unstable Illustrator Anima is a checkpoint made to satisfy anime lovers, the style is similar to professional illustration and has been improved from the previous version stage รNA .
Nsfw is much better, now the model can handle nudity better, the style is similar to Yamer's Anime V5 because it is the main component of the รNA and Anima checkpoint, now using the XL_VAE_C E7 (already baked in) VAE that gives an even better saturation to the image.
This model is perfect for stylized images, abstract art and anime!
Have fun exploring the vastity of styles and images that you can create ^-^
--> REDESTRIBUTION OF THIS MODEL TO PLATFORMS OUTSIDE CIVITAI AND SELLING IMAGES WITHOUT MY WRITTEN CONSENT IS FORBIDDEN! <---
You can contact me on Discord: (yamer_ai), Twitter: @YamerOfficial, email: [email protected] or directly using the Civitai chat system for my approval.
Yamer's Anime version 1 to 5 is a group of models that is specialized in anime like images, this model is being added in the "Ultra Infinity (now called) Unstable Illustrator" family because it follows the same theme (anime), this checkpoint is not specialized in nsfw content, I recommend to use a LoRA for specific images.
When creating this checkpoint I trained it on anime images in various styles, this model is not completely trained, it is a a trained and merged models, other anime models were used as base to achieve the current style that is unique to this model.
(OLD, SD 1.5!) Ultra Infinity V2 - Yamime are models that specializes in anime-like images and sharp results, there is no need for trigger words because the model will already give you anime style results, I advice users to not go beyond a CGF scale of 25 because they colors may break and corrupt, please use the model between 7-25 for sharp colorful results that will satisfy your need without destroying the image.
This model is a big improvement from the previous version with the new sharp style and a better recognition of anime stylized compositions like bodies, background and objects.
Pฬฝอeฬฝอrฬฝอsฬฝอoฬฝอnฬฝอaฬฝอlฬฝอ ฬฝอRฬฝอaฬฝอtฬฝอiฬฝอnฬฝอgฬฝอ:ฬฝอ
Quality: 9/10
Stability: 7.5/10
Flexibility: 7/10
CFG Resistance: 25/30
Style: 9/10
NSFW: 7.5/10
Creator Note: "If you found the results achieved by this model to be satisfying, I would greatly appreciate it if you could kindly consider leaving a positive review. Your feedback and support are invaluable to me." :3
๐ฝ๐๐๐๐๐ค ๐จ๐๐๐ฃ๐ ๐ฅ๐๐๐ค ๐๐ ๐๐๐ ๐๐ค ๐๐๐ฅ๐ฅ๐๐ฃ ๐ฅ๐๐๐ ๐ฃ๐๐๐ฆ๐๐๐ฃ ๐๐ป๐๐๐.๐:
No embedding needed
Better variety of style
The refiner is not needed
Less AI generated look to the image. (When creating anime or realistic images for example)
No face fix needed
Better saturation, overall quality, texture and tonality
Less blurry images (images look sharper)
No adetailer required (Just use hires fix)
เธ ^โข๏ปโข^เธ Works best using:
Restore Face: It is not needed to generate high quality images (Not used in the current showcase.)
Clip skip: 1-2 (The model works well with clip skip set to both.)
SDXL Refiner: Not used with my models
1) Show em big mommy milkies in my dm.
2) A massive tiddie generates a gravitational field by warping the geometry of the surrounding spacetime.
3) Anime hentai is the highest form of research.
-Yamer๐ฝ๐๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐:
Resolution and steps: 1024x1024, 576x1024, 768x1024
Steps: 20 to 50
CFG: 5 to 20
Hires Upscaler: 4x_foolhardy_Remacri or 4xUltraSharp
Hi, I'm an Italian creator on a mission to spread the joy of using AI to generate images and the progress it has made in the past few years. I craft checkpoints and prompts that anyone can use to generate their own art and share it. My ultimate aim is to be a good creator, sharing the models I create as a hobby with others.
Description
New VAE used + better NSFW
FAQ
Comments (11)
The colors are really good looking and the showcased images are excellent!!!
๐ฅ๐ธ
For some reason, I am getting not getting clear eyes like the Super Saiyan Broly example. I am using V5+UAE.
If you haven't already use Hires Fix, download 4xUltraSharp.
Do 25 steps using HiresFix
why it's always generate female characters? I can't generate male characters at all.
If you are having problems generating male characters you can either try V4 or give more weight to the male part of the prompt.
Ex: (male character:1.4)
@Yamerย I tried giving more weight but not worked
I really like the model .. is that v4 deferent style from the last version?
The style is different but you should be getting better males
Hi, I am having an issues with this checkpoint (for Staged รNA). For me it only generates noisy images, I dont know whats the issue.
here is the link to the question
I really love the images generated by this model..
I had a look at your workflow on reddit, I'm not too familiar with some of the nodes you are using, but there are two things I noticed, first your "SDXL Prompt Encoder" node has its base width and height set to 4096, though it seems like you are trying to generate a 1024 by 1024 image and I am not really sure what this node is for (or rather why you cant use the default text encoder). Second, you have "refiner prep steps" set to 0 on your sampler, I think that means that after 0 steps that the refiner model starts working, as in for all the steps you set it to sample, but I could be wrong about this, I don't use the refiner model since most custom SDXL models don't need it. It might also be worth trying to use a simpler workflow to see if the problem is with the model. In any case I hope you are able to get this model working!
@ComradeMittensย Yeah I agree that something was wrong with the workflow, thats not my workflow I got it from civitai, I tried this checkpoint with other workflow (Searge's) and it is working fine.
Details
Files
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.













