MeinaMix objective is to be able to do good art with little prompting.
I have a discord where you can share images, discuss prompt and ask for help. https://discord.gg/meinaverse
我有个可以让你分享图片和参与讨论与询问问题的discord群。
I also have a ko-fi and Patreon page where you can support me or buy me a coffee <3 , it will be very much appreciated:
https://ko-fi.com/meina and https://www.patreon.com/MeinaMix
MeinaMix is officially hosted for online generation in:
- SeaArt
- Mage.space ( with animate feature )
MeinaMix and the other of Meinas will ALWAYS be FREE.
Cover image lora made by: FallenIncursio | Civitai
Recommendations of use:
--------------------------------------------------------------------------------
Enable Quantization in K samplers.
Hires.fix is needed for prompts where the character is far away in order to make decent images, it drastically improve the quality of face and eyes!
---------------------------------------------
Recommended parameters:
Sampler: DPM++ SDE Karras: 20 to 30 steps.
Sampler: DPM++ 2M Karras: 20 to 60 steps.
Sampler: Euler a: 40 to 60 steps.
CFG Scale: 4 to 9.
Resolutions: 512x768, 512x1024 for Portrait!
Resolutions: 768x512, 1024x512, 1536x512 for Landscape!
Hires.fix: R-ESRGAN 4x+Anime6b, with 10 steps at 0.3 up to 0.6 denoising.
Clip Skip: 2.
Negatives: ' (worst quality, low quality), (zombie, interlocked fingers) '
Negatives if you can't use Hires.fix:
'(worst quality:1.6, low quality:1.6), (zombie, sketch, interlocked fingers, comic)'
--------------------------------------------------------------------------------
In the merged models list: MeinaMix V1~11, MeinaPastel V3~6, MeinaHentai V2~5, Night Sky YOZORA Style Model, PastelMix, Facebomb, MeinaAlterV3 i do not have the exact recipe because i did multiple mixings using block weighted merges with multiple settings and kept the better version of each merge.
Description
If you have any questions related to the model or its samples, how to get started or just want to share your creations, feel free to join my discord: https://discord.gg/F6yeQtEQ98 <3
Hires.fix is needed for prompts where the character is far away in order to make decent images, it drastically improve the quality of face and eyes!
---------------------------------------------
Recommended parameters:
Sampler: Euler a: 40 to 60 steps.
Sampler: DPM++ SDE Karras: 30 to 60 steps.
Sampler: DPM++ 2M Karras: 20 to 60 steps.
CFG Scale: 4 to 11.
Resolutions: 512x768, 512x1024 for Portrait!
Resolutions: 768x512, 1024x512, 1536x512 for Landscape!
Hires.fix: R-ESRGAN 4x+Anime6b, with 15 steps at 0.4 denoising.
Clip Skip: 2.
Negatives: ' (worst quality, low quality:1.4), (zombie, sketch, interlocked fingers, comic), '
--------------------------------------------------------------------------------
FAQ
Comments (17)
great model if not for the same one face on all images
It works pretty well with loras that change the style/face though.
Use celebrity faces or LORAs. The dfference between, for example, Morena Baccarin and Dichen Lachmann is noticeable.
what's the difference between v9 and v10?
Thats the hardest questions to answer without using a side by side plot >.<
But basically V10 have better eyes and the style is a bit different thanks to removing some things that were making weird generations
This is a very great model. It's simple to use.
Like this model with little prompt amazing result, I speechless. Many VAE is compatible with this...
Hi, I use Hires. Fix even on normal images (that are not in a distance) and is that a bad thing? The results I get become very blurry and pixelated. This doesn't happen with other models though so just wanted to know if I'm doing something wrong with the settings. Using the Hires.Fix parameters you mentioned in the description
thats... weird o.O are you using one of the latent upscalers by any chance? because they can't have denoising less than 0.55
Are you using face restoration? Sometimes that has this effect for me.
@Meina Hello, can you help me? I wonder how you get these high res images. How can i upscale them? When i use img2img it just adds a lot of other stuff
Hello, @Meina-san. First of all, thank you for the excellent models and your service to the Community.
I keep my personal projects separate between completely Photorealistic and completely NPR/2D work. I never half-ass anything, so I'm not too keen on the thousands of semi-realistic or 2.5D models, though I still like 'em. Your MeinaMix is my primary model for 2D, with @Lykon-san's AnyLoRA as my secondary. AnyLoRA's really good, but as an AI slave-driver (well,kinda,lol), I've already found its limits, I haven't found MeinaMix's. Which brings me to my question: I read in another post that you said that MeinaPastel is supposed to be the "proper" 2D sister of MeinaMix. So, ALL THIS TIME, I wasn't even using the "right" one?? Damn. (BTW, I've only been into the world of AI for a few weeks now, so "all this time" really just means a few seconds in you veterans' perspective.) But regardless, I've been so happy with what MeinaMix has been giving me--EVEN with "Western" things that I tried (like comics stuff,friends couldn't guess it's from an anime model). And yet, even then, you're saying I haven't actually unlocked the Meina family's true 2D specialist. Wow. I'm gonna be doin' a LOT of rework this weekend. (Well, it's gonna be really fun, so it's okay.)
Heh. Anyway, thanks again for everything, author-sensei. Please continue the superb work, have a great week ahead, and cheers. ❤
Hello, what model did you train the checkpoint with?
I want to train a character Lora that match the checkpoint style.
Is it possible to get ckpt or pickle formats? Safetensors blows. I like programming from scratch rather than using automatic1111 or invokeai
sure, will do it to the lastest version of all my models!
@Meina I was wondering why you updated ckpt models, I personally don't use them but I can appreciate having more options! Thanks for the models Meina ^^
@Meina i really appreciate it
Details
Files
meinamix_meinaV10.safetensors
Mirrors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
mei-mx-v10.safetensors
meinamix_meinaV10.safetensors
Meina_V10-baked_VAE.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
Meina V10 - baked VAE.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
Meina V10 - baked VAE.safetensors
兼容漫画meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
meinamix_meinaV10.safetensors
Meina V10 - baked VAE.safetensors
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.







