You can use my model directly at : Tensort.Art | Mage.Space
_____
On top of that, SDXL is only in its early stages, I may have many shortcomings. I hope you can give me suggestions for corrections in the next updates. If you find it useful, please vote 5 stars and post an image to help me, it's really important, I thank you all very much
_____
📖 Checkpoint Info
Info : This is a training model based on the best quality photos created from SDVN3-RealArt model. With the desire to bring the beauty of SD1.5 to SDXL model. I added a bit of real life and skin detailing to improve facial detail. Currently I have two versions Beautyface and Slimface.Detailface version is an upgrade of Beautyface, it creates more skin details and creates better realism than the beautyface version. With the Beautyface version - the model face will have more diverse angular lines. In contrast, the Slimface version will have a slimmer and easier-to-see model face, suitable for Asian faces.
_____
You can refer to some of the indicators below to achieve the best image quality :
Steps : > 50
CFG : 9-10
Size : 768x1162 px ( or 800x1200px )
You can also use hiresfix ( hiresfix is not really good at SDXL, if you use it please consider denoising streng 0.2-0.3 ) or After Detailer
Sampling : Euler a or DPM ++ SDE Karass
📝 Quality Prompt :
Base quality prompt :
(masterpiece,best quality, ultra realistic,32k,RAW photo,detail skin, 8k uhd, dslr,high quality, film grain:1.5),📝 Negative Prompt :
(worst quality, low quality, illustration, 3d, 2d, painting, cartoons, sketch), tooth, open mouth,💡 Concept reference:
Portrait Lady Real :
(masterpiece,best quality, ultra realistic,32k,RAW photo,detail skin, 8k uhd, dslr,high quality, film grain:1.5),
{short hair | long hair}, __*/haircolor__ hair,lady, brown eyes, __*/colors__ __*/Wear/*__, freckles,
(__*/colors__ gradient background, simple background:1.3)_____
Also you can test those prompt tags quickly with my model here : Tensor.art
I have written a colab site that integrates all tools for you to use stablediffusion without configuring your computer, you can refer to : Colab SDVN
I have provided an album with more than 2000 original photos with full prompt, you can refer to understand how to use SDXL : SDVN-SDXLPrompt Kit
In addition, some of my models are available on the Mage.space platform, you can refer to: SDVN Mage
_____
On top of that, SDXL is only in its early stages, I may have many shortcomings. I hope you can give me suggestions for corrections in the next updates. If you find it useful, please vote 5 stars and post an image to help me, it's really important, I thank you all very much
_____
👨🏻💻 Creator : stablediffusion.vn
📢 Discord : Discord SDVN
📢 Facebook : Group SDVN
☕️ Donate me : Bank | Buymeacoffee
Description
FAQ
Comments (16)
Cool model, but concerned about the idea of training a model on output from SD 1.5 based models. It might learn the artifacts that SDXL could avoid, just something I think people should be aware of especially if someone wants to make a mix model
Thank you for your feedback, I have tried to filter the images very carefully to avoid errors as much as possible
This non-real "hyper plastic look" is the balm I have been searching for.
Photo realistic SD1.5 models are OK I guess, but the Barbi-smooth doll-esq features SDXL promises, not to mention, the bold move towards a bright sexless, non-erotic future has me, riveted, to the screen.
Kindest possible regards to the makers of SDXL and the future they promise us all!
I’m just grateful that the makers of SDXL have agreed that erotica is Overrated and G-rated output is the way to go.
bro, go look at Hassanblend V1, it'll all be okay. What are you uploading?
I have tried several SDXL models, even those from more famous modelers, but none are as stable as this one. You should give it a try
Thanks buzi for the feedback, I will update and improve it more soon
There are only two kinds of checkpoints: SDVN and the rest. ♥️
thanks you so much
Same problem like all the other trained xl models there is no corresponding trained refiner. My summary: I have tested the trained xl models published recently on civi and the native xl base model with several different workflows (With, without refiner, refiner sd1.5, base and refiner xl with upsclae sd.1.5. Conclusion: xl base and refiner xl are useable for some outstanding outputs. But the trained sd.1.5 models are superior to their xl variants and my opinion switching to trained xl models makes only sence if there is also an appropriate trained refiner available because the standard 1.0 refiner doesnt fit to the trained xl models. Just compare ure results gen by SD1.5 models and the corresponding xl model and u will know what I am talking about.
I understand what you say, I am also testing and have options to train and fix errors, trying to fix what you say. Right now, I'm doing it. Thank you for giving me feedback
I am not using refiners at all. Just use the Hi-Res fix (I also pump up the resolution to 1280x1280) at 0.2 denoise and it does an amazing job.
@Daedalus_7 yes, for hires fix at sizes larger than 1600, i find that denoising strang 0.2 is the best
So for these trained models you guys are not using the refiner, you're just using a 1.25 Hi-Res fix with a small amount of denoise? I didn't try such a small upscale but I think I tried 1.5 and got OOM errors so I gave up (and I'm on an Asus 4090). Thanks, will give this a try today.
I have been using 1.5 models as a refiner for now, it allows the creativity of SDXL base with the quality of 1.5 models, especially with fine details like faces and skin texture.
@aadsgdgadadggda Currently, I train this model based on the method of combining the base image from 1.5 and the real image to add skin details.
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.