I turned myself into a Disney / Pixar Character - Stable Diffusion Image to Image Tutorial
HTML-код
- Опубликовано: 9 окт 2022
- Support my Channel:
/ @oliviosarikas
Facebook AI Group: / theairevolution
How to Deepfake: • I faked JENNIFER LAWRE...
Stable Diffusion Install Guide: • Stable Diffusion Insta...
Automatic 1111
Subscribe to my Newsletter for FREE: My Newsletter: oliviotutorials.podia.com/new...
How to get started with Midjourney: • Midjourney AI - FIRST ...
Midjourney Settings explained: • Midjourney Settings Ex...
Best Midjourney Resources: • 😍 Midjourney BEST Reso...
Make better Midjourney Prompts: • Make BETTER Prompts - ...
My Facebook PHOTOGRAPHY group: / oliviotutorials.superfan
My Affinity Photo Creative Packs: gumroad.com/sarikasat
My Patreon Page: / sarikas
All my Social Media Accounts: linktr.ee/oliviotutorials - Хобби
Support my Channel:
ruclips.net/channel/UCCKx8mAHiFus-XYQLy_WnaAjoin
Facebook AI Group: facebook.com/groups/theairevolution
How to Deepfake: ruclips.net/video/kJEHJXiwGVU/видео.html
Stable Diffusion Install Guide: ruclips.net/video/Pyze0seDHzA/видео.html
Such a good tutorial! I like how you explain what each setting does and not just changing it!
Detailed and scrupulous; I liked the video very much and it is invaluable in being able to understand sd in these early stages. Thank you!!
Your videos are really helpfull and creative, thank you !
Very userful, thank you!
Thanks for your knowledge Sir
Fn great information here thank you. Denoising was my issues. Commented and SUB!
From my experience I'm not believing the "you only need 40 steps for euler" that a few people keep saying. Especially for big prompts using compositional prompting. 50-55 is the minimum I tend to use
same for me, the coolest and most photoreal images ive gotten were made with 50-55
Euler-a with more than 30 steps is just creating new stuff not better stuff. Try it step by step...
well, it kind of depends on what you do. You can always try more. For example for outpainting, you want to use a lot more steps, like 60-100 even with euler a
cool video , thank for sharing i would love to see more content like this super fun
thank you :) If you have any suggestions or challenges, let me know
Cool video!
I apparently chose a different way to install SD locally and my UI looks totally different and lacks the cool features like "interrogate". My UI seems to be from a "cmdr2, ko-fi". can I exchange it to the more powerful UI?
Olivio...I am a straight guy but would give you a kiss on your hairy face for all the tutorials that you have done.
Thank you for sharing all these wonderful tutorial. Clear, and precise presentation.
😚😚😚😚😚👄
As mentioned by others, I too was getting realistic results... anything Pixar like was not similar to the source image and vice versa. I am getting much better results with the `nitrosocke/mo-di-diffusion` model (and `modern disney style` in the prompt)
How do you add the model? :)
Hi , olivio . is there any way to auto send prompts to midjourney discord , I tried to send auto messages with /imagen prompt: blue but the midjourney bot didn't took action , do you suggest way to automate prompts ?
Hi May I ask a question? Can I try Stable Diffusion Offline in my own pc?
great video
the animation at the end give u a like xD
This is the DaVinci Resolve LUT girl!
please give me a link to it, this is my first time learning about it, sorry if i did something wrong. thank you
ty
So when you use the prompts as well as an image in img2img, is Stable Diffusion using data from both the text prompt AND the image you've given it? Sometimes generated images seem really far off from the original.
yes, it's a lot of Experimentator. The denoise slider defines how much of the original image is used. 0 i all the image 1 is non of the image
What software do you use?
interrogate clip won't work in my PC, because Out of CUDA Memory, dreambooth and preprocessing model don't work too, the same error
Hey, thank you for the tutorial.
But when I'm pressing on Interrogate it only says "Loading..." above the prompt field, without ever finishing the loading process. Do you perhaps know of a fix for this?
Hmm nevermind, after restating it again it filled out the prompt. But it ended it with
@@TryHarderJ look in your CMD window to see the error message and ask for that error in the official Stable Diffusion Discord. they are often very helpful
@@OlivioSarikas Thank you, will do! :)
How can i fix this error > TypeError: 'NoneType' object is not subscriptable
Which model does he use?
Hi is it possible to create coloring book page using stable diffusion?
Yes.
Yes, that should be possible. All it requires you to find best prompt with best configs.
@@rajeshkannanmj stuff like "simple" "outline" I'm sure there are even specific styles too apply
whats the difference with people adding 4k, 8k, 16k, 32k to their prompts?
can this be done with DiffusionBee too or do you suggest to go with Automatic1111 ?
same question here. I have an iMac and downloaded DiffusionBee and I don't get all the features that he's showing in the video. Also, any other models I can download to help get the disney like characters?
It didn't work for me. I need your help! ---> RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory
is there an easier method...? That's too much downloading things and I stuck.. :-(
I can't see restore faces anywhere?
I'm having this problem with the interrogation part, "RuntimeError: CUDA out of memory, Tried to allocate 340.00 MiB" it actually generated some promt words halfway but at the end of line it says
try downscaling the image to a lower resolution
hello could you recommend an artificial intelligence that could generate an avatar for me and in addition on which I could integrate one of my images
What computer spec you need for this stable diffusion..rtx 3060 enough for the gpu?
I run it with a RTX2060.
Successfully.
I run with a GTX 1050 Ti just fine it takes like 20 seconds to generate
@@PawFromTheBroons my current setup is gtx970 only :D..n ram 8gb..i want to try run it and see whether can generate a pic
@@namae- wow 20 second only?i will try soon with my gtx970 hehe
Ok I’m lost here! Where should I get this software? If this a free software?
I literally point you in this video to a install guide in the first seconds of the video ;)
grazy, mein herr. was es alles gibt
crazy welt in der wir leben :)
I don't have interrogate on mine. Did I miss an install feature? Where would I find that and download it? Can anybody help me?
same problem, smbd help
I just wish midjourney had something like this.
yea but its cool they are both doing their thing. enjoy them both
where are link ?
in the video description and the pinned comment
Wonder why ai can't draw hands?
LOL. And wow.
Is that Tom Segura?
Nice work. I wouldn't associate myself with Disney though.
For me keeps showing realistic results, not disney character look
This did not help me, probably because it's outdated. It would be very helpful if I were seeing on my computer what I'm seeing in the video.
First comment 😎❤️😎
welcome! Even faster than my sticky comment! 👍😀
@@OlivioSarikas 🤣🤣
👍👍
wrr
Haha, you could have at least not just used a default character from Character Animator.
Bro, I am the default character ;)
Are we ever going to get past this AI obsession?
good idea, but unf. I got this error "RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.cuda.HalfTensor) should be the same" ...
Why do you keep saying X, Epsilon? It's a Y, as in XYZ.
Because he is from Germany and there the letter Y is called Ypsilon, not Wye