Stable Diffusion - Better Images with FreeU - Fooocus

Подписаться 12 тыс.

Просмотров 10 тыс.

50% 1

Master Generative AI for Creative Image Generation! 🎨 In this video I check out the secret hidden FreeU:Free Lunch in Diffusion U-Net settings in Fooocus. #Fooocus #StableDiffusion #DigitalArt #ImageManipulation #CreativeTools
Fooocus is a great interface for using Stable Diffusion and is very easy to get started. It also will run on systems with only 4GB VRAM.
I am running on an i5 32GB RAM Nvidia 3070 8GB VRAM
If you find the video useful and would like to tip, RU-vid does have the Super Thanks option. Those are greatly appreciated so I can have my coffee!!
☕ BUY ME A COFFEE: ko-fi.com/kleebztech
Download Fooocus: github.com/lllyasviel/Fooocus
How to install Fooocus video: • Stable Diffusion - Dow...
Fooocus Beginners Guide: • Super Easy Stable Diff...
Fooocus Upscales and Variations: • Stable Diffusion - Foo...
Checkpoints and LoRAs the Basics: • Stable Diffusion - Che...
Fooocus Poses and FaceSwap - Image Prompts: • Stable Diffusion - Pos...
Zoom with Outpainting: • Stable Diffusion - Zoo...
Inpainting with Fooocus - Don't Regenerate, Fix!: • Stable Diffusion - Inp...
Background Replacement with Fooocus: • Stable Diffusion - Bac...
Fooocus Tips and Tricks (and AI Hands) • Overlooked Fooocus Tip...
Stable Diffusion - Styles in Fooocus - How do they work?: • Stable Diffusion - Sty...
FaceSwap and Consistent Character TipsPart 1: • Stable Diffusion - Fac...
FaceSwap and Consistent Character Tips - Part 2: • Stable Diffusion - Fac...
00:00 Intro
01:09 FreeU settings and what it does
02:45 Comparison Images
06:30 Best way to use

Наука

Опубликовано:

11 июл 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 46

@KLEEBZTECH 3 месяца назад

If you find the video useful and would like to tip you can buy me some electricity for all these image generations. It is greatly appreciated! ko-fi.com/kleebztech

@romanmakarov2021 2 месяца назад

SO Frecking useful!! Thank you Kleebz! Your videos are always great, unique and valuable=)

@mick7727 3 месяца назад

free lunch! What I love most about this stuff is there's so many different ways to go about generation. I've wasted days upon days just creating something that popped into my head, just because I felt like I wanted to see a little more. I'm glad I found your channel, sir, it really helped me get into Fooocus and now I'm hooked.

@KLEEBZTECH 3 месяца назад

I can totally relate to that. I'm glad you enjoy the content. I end up not liking it because I end up watching it too much during editing. 😂

@TheColonelJJ 3 месяца назад

Thanks. I had no idea what FreeU was used for much less capable of. Good Turorial!

@KLEEBZTECH 3 месяца назад

You're welcome! I still don't fully know but I at least now have a little bit of an understanding myself. Hard to do extensive testing and having to adjust it constantly. But I really think it's something that you need to experiment with to get a feel for what does what.

@b.s.design 3 месяца назад

Great workshop! Thank You!

@petrimannisto9318 Месяц назад

My thoughts exactly. It's just a smoke without fire claim to say that FreeU makes your generations always better. It's very subjective. I've found it most useful to fix a finger or two or pose in an image without having to inpaint, when you wiggle around with FreeU settings.

@hmmmmmm_3429 3 месяца назад

freeU basically plays with VAE layers of model, in sd1.5 to sdxl there are 4 VAE channels it basically plays with the weights of those channels and there is no telling which channel is better or contains what its mostly used to correct color. also SD3 will have 16 VAE channels..

@AIEnthusiasm 3 месяца назад

great information

@labkarmonsolutions956 3 месяца назад

Thank, teacher =)

@KLEEBZTECH 3 месяца назад

You are welcome!

@mik3lang3lo 3 месяца назад

Great video

@KLEEBZTECH 3 месяца назад

Thanks!

@sheldonventer2849 3 месяца назад

Great vid :)

@KLEEBZTECH 3 месяца назад

Thanks!

@lilillllii246 3 месяца назад

Thank you. It's a little different question, but how can I use the eular a sampler in fooocus?

@southcoastinventors6583 3 месяца назад

Did it have any effect on accuracy like the Viking ship with FreeU enabled was better than without because it fixed the mast placement or maybe better fingers ? Thanks for the setting guide hope we see SD3 come out this month.

@KLEEBZTECH 3 месяца назад

For some images yes. But I haven't run enough images to really determine much about that.. It could just be that it places them differently and not necessarily better because of certain settings. If I figure a way of automating it then I might have better chances of testing. Just a real pain to run tons of images this way. I might be able to do that in something else though.

@Designing-hc5pz 3 месяца назад

thanks a lot for this can you make a tutorial on how to take piece of cloth and by doing image prompt on this we can dressed it to any person/model i request you to do this & keep going 💪

@KLEEBZTECH 3 месяца назад

I would love to do something like that but I don't know of any real way of doing that with the image prompt reliably. Simple clothing you might get away with it for the most part but not for anything more complex. I'll play around with it some more and see if I can find anything to make it more reliable.

@Designing-hc5pz 3 месяца назад

@@KLEEBZTECH i've got an similar tutorial for you if you're justgetting started ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-U8vyw1nbJlM.htmlsi=9p-tA583-YeilKno

@trabajoxmoney6857 3 месяца назад

Muy buen video, como siempre , aportas gran cantidad de información muy valiosa. Tengo una duda , haber si me puedes ayudar , que tiempo te tarda a ti en generar 1 o varias imagenes con SDXL en fooocus o otras herramientas con tu rtx 3070? , yo tengo actualmente un rayzen 7 1300X con una GTX 1070 y no se si valdría la pena instalar fooocus en local o mejor generar las imagenes en google colab. Espero tu respuesta , muchas gracias por el video. Un saludo

@KLEEBZTECH 3 месяца назад

It takes about 35 to 40 seconds. If doing locally you may want to do the lightning ones. Yours will be very slow.

@julianopajaro 3 месяца назад

Hi! Thank you for sharing with us this Fooocus tutorials. I really appreciate. I have a question. Is it possible to increase the number of the steps in Fooocus? For example: Quality goes to 60 steps, can I make the quality goes to 100 steps?

@KLEEBZTECH 3 месяца назад

I think you can only do that in the debug settings with the "Forced Overwrite of Sampling Step" slider. Drag that up to 100.

@julianopajaro 3 месяца назад

@@KLEEBZTECH You got it! Thanks!!!

@voxyloids8723 2 месяца назад

Could FreeU help to create consistence batch for a video?

@KLEEBZTECH 2 месяца назад

I'm not exactly sure what you mean but probably not.

@heinzwillikoch2715 3 месяца назад

Thanks.But we can I install juggernautXL8 . I have the juggernautXL -6 Version.Can you help me?

@KLEEBZTECH 3 месяца назад

This video should get you in the right direction. You'll just download that checkpoint and put it into the folder. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-GgEEb2K0j7s.html

@heinzwillikoch2715 3 месяца назад

I found it.Nr.8 works well.Can I remove juggernaut 6 ? Thanks

@KLEEBZTECH 3 месяца назад

You can but depending on how your preset is set up it may try to download it again if it doesn't detect it. Give it a try and find out.

@AK47_64 3 месяца назад

Just wanna ask why do you need image generation! What do you do with them?

@KLEEBZTECH 3 месяца назад

Some of it is just I like to create stuff. I also like to learn. So I figure as I learn I might as well share what I learn. But I have gotten some shirts done and prints made up for the wall. And some is work-related.

@KLEEBZTECH 3 месяца назад

Plus I have a couple future projects I want to do with this stuff but AI isn't quite where I need it to be for what I want.

@AK47_64 3 месяца назад

@@KLEEBZTECHI also have some future projects, such as content creation through AI-generated videos. However, the text-to-video capability is not quite capable enough yet.

@KLEEBZTECH 3 месяца назад

And I love creating my own desktop backgrounds. I have several monitors and I have the backgrounds cycle through different ones.

@KLEEBZTECH 3 месяца назад

The key is to be ready when the capability becomes available. Plus I love experimenting with everything. So many options.

@techendofficial 3 месяца назад

Is juggernaut xl v9 better or juggernaut xl v8......??

@KLEEBZTECH 3 месяца назад

I don't know if either one is better. I use nine occasionally but still mostly eight.

@Infiniteinsights0 2 месяца назад

Can foocus do Img2img batch

@KLEEBZTECH 2 месяца назад

No.

@ThisClayman 3 месяца назад

i see people using juggernaut v8 and not v9, you know why is that?

@KLEEBZTECH 3 месяца назад

Most people are still using the default. Plus I find just because it's a larger number doesn't necessarily mean it's better. Often these are trained a little bit differently. And when people get used to a certain model and they switch they can end up with different results.