A-Zovya Photoreal - CivArchive (CivitAI Archive)

A-Zovya Photoreal - V2

NSFW

Update June 28th, added V3+VAE, pruned and fp16 versions also added.

V3

More training added to supplement some things I feel are missing in current models. Skin textures, darker lighting and more variety in race in people.

V2

After weeks in the making, I have a much improved model. No longer a merge, but additional training added to supplement some things I feel are missing in current models. Lots of new training for skin textures, lighting and non-asian faces to balance out the asian dominance in models. If you create a generic prompt, you'll get a greater variety of races and faces now. Skin textures are increased by a large amount, if that's not your thing, you can put detailed skin in the negative prompt and get back that airbrushed look if you like.

A photorealistic model I've been using for a little bit. I hate smooth airbrushed skin so I refined this model to be very realistic with great skin texture and details. There are 2 models, a standard and an ultra.

The Ultra model is nearly 3 times as large, but it's not 3 times as good. It is better though with a broader knowledge. You can even generate 1024x1024 in many cases. The config file is required for the Ultra model, be sure to download that too.

So I recommend using the normal version unless you have the need or vram to run the Ultra model. If you'd like to run the Ultra model with modest vram, try --medvram or --lowvram in your auto1111 startup script.

A side-by-side comparison. Link to see full-size samples with metadata. Like I said, the Ultra model isn't 3 times as good, but it is better across multiple subjects, not all. The pruned version of the Ultra model is debatable, might be better to use the normal V1 model in some cases. But all have been provided so you can make the choice yourself. For me, I use the Ultra model in every case.

Do you have requests? I've been putting in many more hours lately with this. That's my problem, not yours. But if you'd like to tip me, buy me a beer. Beer encourages me to ignore work and make AI models instead. Tip and make a request. I'll give it a shot if I can. Here at Ko-Fi

Description

Stable Diffusion 1.5 (512 x 512)
A-Zovya Photoreal V2
VAE recommended vae-ft-mse-840000-ema-pruned.ckpt

FAQ

Comments (40)

1603456Jun 19, 2023· 5 reactions

CivitAI

Sad i can't give more than 5 star, nice work.

Zovya

Author

Jun 19, 2023· 1 reaction

thanks!

1603456Jun 19, 2023· 1 reaction

@Zovya I work for long time to make new one but still not good, Trained more than 800k image's and still not good, what i have here just marge for my fragment work.

Zovya

Author

Jun 20, 2023

@BoRnNo0b oh, sad news. I think the best route is training and merging. Train to fix what you don't like.

1603456Jun 20, 2023· 1 reaction

@Zovya I'll try, thanks for reply mate.

JazzyTechJun 20, 2023· 1 reaction

CivitAI

Currently throwing prompts at V2 and they are looking mighty fine so far :)

MasmalkarJun 20, 2023· 8 reactions

CivitAI

My poor storage, but soooo many good checkpoints lately, thanks for your model man

SnG17Jun 20, 2023

My SSD is in despair as well.

frfromgJun 20, 2023

Wait for smaller version (hopefully).

Zovya

Author

Jun 20, 2023· 2 reactions

I'll get a pruned version up tomorrow.

1845798Jun 20, 2023· 2 reactions

CivitAI

She's not just a woman, she's a goddess.

bullseyetrollJun 20, 2023· 2 reactions

CivitAI

I thought realistic model has hit a bottleneck.
I was wrong

Ant6431Jun 20, 2023· 8 reactions

CivitAI

Pruned version pleease

smoonHackerJun 20, 2023· 1 reaction

yes! Prune it!!

nudilliJun 20, 2023· 5 reactions

CivitAI

Great work. Waiting for the inpainting v2!

SoochJun 20, 2023· 1 reaction

CivitAI

Happy to see a V2! Will there be an Ultra version of this one?

Zovya

Author

Jun 20, 2023· 1 reaction

No, I created this one in a more traditional way. You could merge the two though.

JazzyTechJun 20, 2023

Sigh, I asked this in another message. I was wondering the same thing. ;)

7727Jun 20, 2023· 2 reactions

CivitAI

Very excited to try this one, whats the VAE used here or is it baked in? Or is it the usual 84000

Zovya

Author

Jun 20, 2023· 1 reaction

Thanks, yeah it's the usual 840000

LostAtSeaJun 28, 2023

@Zovya version 2 is a little fat can you upload the trimmed version

Zovya

Author

Jun 28, 2023· 1 reaction

@LostAtSea pruned version added

open_promptJun 20, 2023· 2 reactions

CivitAI

Wow, this is amazing!

TheLoraCollectiveJun 20, 2023· 3 reactions

CivitAI

What is the difference between a model and an inpainting model? I can't find a clear explanation.

aashiJun 22, 2023

same

settima_aiJun 22, 2023· 8 reactions

@aashishvishwakar7422 The inpaint model is trained to do the inpaint not just replacing the area with a new version, but generating a new detaill compatible with the existing image. In other words, is 100 times better than the inpaint done with another generic checkpoint!

settima_aiJun 22, 2023· 1 reaction

@aashishvishwakar7422 A clear example can be see here:
https://preview.redd.it/how-to-turn-any-model-into-an-inpainting-model-v0-bb3ydsincy8a1.png?width=3928&format=png&auto=webp&v=enabled&s=c7e791aca3bee37d6c1d0a6d2908f22876231bd7

jcc021058Jun 24, 2023

@settima_ai so, you main inpainting model not need after edit ?

settima_aiJun 24, 2023

@jcc021058 Not clear your question. Anyway, you need the 'inpaint' model only if you want to try some inpaint after generating an image. Otherwise you dont need it!
(also, if you dont know what 'inpaint' is, you definetely dont need it!)

fifthdreamJun 20, 2023· 3 reactions

CivitAI

I just want to thank you for this model. I'm now achieving incredible results with a specific subject/style that i've been working on a long time. It's really amazing. I'm so impressed and happy with this model.

loades111735Jun 20, 2023· 1 reaction

CivitAI

Am I the only one who gets weird splashes of color on the faces? If not, did you find a way to fix it? Edit: it was an issue with the dynamic thresholding extension.

loades111735Jun 20, 2023· 1 reaction

Seemed to be an issue with the dynamic thresholding extension

smoonHackerJun 20, 2023· 1 reaction

No, you are not the only one

ulicoconut483Jun 28, 2023

I have the same issue sometimes

Zovya

Author

Jun 28, 2023· 1 reaction

My theory is it has to do with the skin textures. At low stages when the resolution is low, it causes artifacts trying to increase the details on faces. Seems to happen in high detail areas like the lips, eyes, etc. If you want the detailed skin textures, the best bet is to cherry pick, hires fix, or inpainting. If not, add (detailed skin) to the negative prompt.

hawkers0Jun 23, 2023· 2 reactions

CivitAI

v1 was one of my favorites! Thanks for the great work!

Btw, will there be an Ultra version for v2 as well?

Zovya

Author

Jun 23, 2023

no. I used a more traditional method for making this model. but nothing stopping you merging it into the v1 ultra.

roynooneJun 23, 2023· 7 reactions

CivitAI

Love your A-Zovya Photoreal model! Waiting for new updates :)

0l1v1aR0551Jun 25, 2023

also, the inpainting V2 model - please!