ToonYou - Beta 6 is up!
Silly, stylish, and.. kind of cute? 😅
A bit of detail with a cartoony feel, it keeps getting better!
With your support, ToonYou has come this far, Thx!
⬇Please read the information below🙏
1️⃣Recommended Settings
- VAE is included starting with Alpha2 (840000)
- Clip skip: 2
- CFG scale: 8- Sampler: DPM++ SDE Karras
Sampling Steps : 30+- Upscaler (Hires. fix): R-ESRGAN 4x+ Anime6B
Hires steps: 14
Denoising strength: 0.35
Upscale by: 1.5+- Default prompts
Pos: (best quality, masterpiece), ...
Neg: (worst quality, low quality, letterboxed), ...2️⃣Using Adetailer
- Main UI > Extensions > Install from URL
This fixes problems with the character's eyes and face in most situations
The default usage is to simply enable it and leave it alone✅Now you're good to go!
For more information, please continue below
ℹ️Why is my image different from yours?
I can't give you any technical help, sorry
Keep your prompts simple and correct
Tag auto complete extension to find the appropriate tags
Tags like 'Realistic, 8k, ...' sometimes ruin an image
Also These tags are often more useless than you might think
Neg Tags like 'mutated arms' are useless than you might think
Neg embedding can have a significant impact on the characteristics of a model
These are not as perfect as you might think
💝Contact & Support
⚠️Warning
You are solely responsible for any legal liability resulting from unethical use of this model
Don't use my model as bait to drive people to your paid generation service without permission
For anything other than general personal use, please be sure to contact me
Description
FAQ
Comments (16)
I love this style so much, but what is the difference between alpha2 and 1? Are they completely different or just an upgrade? I just downloaded alpha1.
Compared to alpha 1, alpha 2 loses some of the contrast and strong colors, but the faces are more defined and the outlines are cleaner. Also, the default characters are slightly taller (proportions have changed).
Personally, I think this is a matter of taste.
Thank you for the model, I'm having trouble loading it, after I select it and press generate, it somehow goes back to whatever previous model I have, do you have any idea why, I would love to try it, Thank you.
I don't know, sorry. Take a look at the error log.
this may be my new absolute favorite model
IDK why but, it only gives low res images with a lot of artifacts and blurry parts. No matter what, every image has some trouble.
I've found this happens more with smaller resolutions on many anime models (like 512x768), it helps to generate larger sizes and having hi-res fix and an upscaler turned on.
@Fenn I'm actually working with 768x768, 768x1024 and 1024x1024. But I'll try other resolutions. Thanks! It's a really cool model btw.
@ElPerroRetro Since it's based on SD 1.5, there's room for it to cause issues at the resolution you set. Increase the resolution using Hi-res
@Bradcatt It's impossible. If I try with a 3060 TI (8GB) it says this: RuntimeError: Not enough memory, use lower resolution (max approx. 1216x1216). Need: 2.8GB free, Have:1.3GB free
If I try with a 3080 TI (12GB), it's the same. So... I asked a friend who has a 3090 and it worked. Sort of. If he tries to make a 768x1024 img using DPM++SDE Karras, SD collapses and he gets the same error; "Not enough memory". How much VRAM do we need to make it work?
@ElPerroRetro I'd suggest just using DPM++ 2M Karras or Euler A, especially for this sort of model.
You could try generating 768x1152, but you need to have Hi-Res Fix turned on so that it generates at 512x768 on the first pass, and then builds the 512x768 into the higher res result.
Not upscale, probably leave upscaler turned off, at least for testing.
@ElPerroRetro At 512x768, hires upscale x1.5 (=768x1152) consumes about 8400mb of vram. These are the settings I typically use, which means this model is usable as intended on a VGA with 10GB of vram.
@Bradcatt Oh. I see. Thanks... Why am I so poor? I should not have been born in Latin America :(
@ElPerroRetro I know other upscaling methods eat less vram, but I'm not sure about this one. The 3060TI is still a great, above average VGA and you're not behind, it's just that the demands of AI image generation are too high 🥲
@ElPerroRetro Oh, I forgot to mention, applying --medvram would help.
@Bradcatt That worked PERFECT. I can even use 2.1 models that I couldn't before. Thanks a lot!
Details
Files
toonyou_alpha2.safetensors
Mirrors
BeenYou_Toon_A2.safetensors
beenyouToon_alpha2.safetensors
toonyou_alpha2.safetensors
beenyouToon_alpha2.safetensors
toonyou_alpha2.safetensors
toonyou_alpha2.safetensors
beenyouToon_alpha2.safetensors
beenyouToon_alpha2.safetensors
pcia_beenyou_toon_alpha2.safetensors
beenyouToon_alpha2.safetensors
beenyouToon_alpha2.safetensors