No video :(

Fooocus Update 2.2.0 - Embedded Metadata | Prompt Array & More! - Stable Diffusion

Подписаться 13 тыс.

Просмотров 10 тыс.

50% 1

In this video I go over many of the new features added to Fooocus including array support and metadata added to images to be more like ComfyUI and others. #Fooocus #StableDiffusion #DigitalArt #ImageManipulation #CreativeTools #midjourney #Dall-e
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
Also I am a verified creator with Brave and do accept BAT for users of the Brave browser.
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
00:45 New File Formats
01:50 Multiple Model Folders and Lora Options
03:50 Array Support
09:00 Metadata Image Support Added

Опубликовано:

5 авг 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 104

@mrbabyhugh 5 месяцев назад

Thanks much for this! I just want to say again, Fooocus is the best thing I have seen made thus far. If they were to add even lip sync, this would blow the others like Roop & FaceFusion to dust. But without video support too, Fooocus is so darn good!

@Enomyx 5 месяцев назад

I hope they'll ad SD1.5 support in the near future. Great coverage of the updates!

@easymad1918 5 месяцев назад

Fantastic.. tomorrow morning ibwill have a try, thx for the hands up

@KLEEBZTECH 5 месяцев назад

Have fun! I love a couple of the new features like array support combined with keeping the same seed.

@alpenjager3986 5 месяцев назад

Great video. Your channel consistently provides the latest information regarding Fooocus

@KLEEBZTECH 5 месяцев назад

Thanks! I will try to keep that up.

@alpenjager3986 5 месяцев назад

@@KLEEBZTECH Can you make a video about tips and tricks using the refiner in Fooocus because sometimes the image generated using the main checkpoint can't achieve a realistic look? In my case, I prefer using the SD checkpoint for the refiner rather than SDXL to achieve the most realistic look because I can't achieve the same result with the XL checkpoint refiner.

@KLEEBZTECH 5 месяцев назад

I will keep it in mind as a possibility. Guess I will have to do some more experimenting with that myself.

@t.n.t.electric1500 5 месяцев назад

Thanks man. Nobody else on RU-vid has explained checkpoints 👍👍👍👍

@KLEEBZTECH 5 месяцев назад

Happy to help!

@TheMaxvin 5 месяцев назад

Wow, Fooocus heard my request! Super.

@AbdulKhamis 5 месяцев назад

Great updates, and a nice presentation of it.

@KLEEBZTECH 5 месяцев назад

Thank you kindly

@clangsison 5 месяцев назад

very detailed, thank you.

@KLEEBZTECH 5 месяцев назад

You are welcome!

@Superficialgallery 5 месяцев назад

Always great, man. Always great.

@KLEEBZTECH 5 месяцев назад

Thanks!

@WhySoBroke 5 месяцев назад

Great updates but I wish they had direct support for SD1.5 models, the refiner method is far from ideal, generated images never fully capture the refiner.

@eugenmalatov5470 5 месяцев назад

thx ))

@MindMotiveMerge 5 месяцев назад

Thx :)

@KLEEBZTECH 5 месяцев назад

You're welcome!

@wernerblahota6055 5 месяцев назад

👍👍👍👍

@knightride9635 5 месяцев назад

It was a good idea to add webp format

@KLEEBZTECH 5 месяцев назад

I am a PNG fan but like the additional options. 👍

@SummarizeAnything 4 месяца назад

Great overview of the latest Fooocus update! I am also use Fooocus to create my videos, you can check them out, if you want. It seems like there are some significant changes, especially with the ability to customize output formats and manage LoRAs more efficiently. I'm curious, though, how do you think these changes will impact your workflow compared to the previous version?

@KLEEBZTECH 4 месяца назад

Well I've been waiting for the array feature and the ability to add metadata to the images. For testing the array feature will be nice combined with using the same seed.

@SummarizeAnything 4 месяца назад

@@KLEEBZTECHhm, i am already have metadata in my fooocus. But i don't know how it can be useful in any way.

@KLEEBZTECH 4 месяца назад

I like the metadata because I can move the files and always go back to get the settings that created it.

@saberkz 5 месяцев назад

Please add shortcut or option to use lightining and turbo models easily

@KLEEBZTECH 5 месяцев назад

I think that may be coming soon. But you could set up a preset for those. I am going to do that soon. Just have not gotten to doing it yet.

@antikoerper256 5 месяцев назад

Hey Rodney ive been meaning to ask you - what do you suggest is better at this stage for a person who doesnt have a PC to run AI imaging - to invest in a laptop or desktop with a modest 8gb video card, to get something in the range of a machine with 12-16 gb VRAM and components and card, or to go for an expensive machine with 24gb or more? Im asking because Im thinking which one is better. Also do you think Fooocus will one day do text to video and how much VRAM would text to video require in your opinion as an entry / base level point? This is what I have to deal with and answer myself these questions before I figure out what to buy and which way to go. Thanks in advance.

@KLEEBZTECH 5 месяцев назад

As for video, I think it is too soon to tell if good video will be able to be done easily on a personal computer. In the near future. I don't think video will be part of Fooocus since that could overcomplicate things and the goal is simplicity. But I could be wrong. As for what type of system to buy.... obviously the best one is going to last you longer and be much better but I'm not sure I would go that route myself yet. I think in another year or two the hardware front will change somewhat and there will be much better Hardware for this stuff. So I don't know if it makes more sense to buy a cheaper system for now and wait for something much better. Although I would try for at least 10 GB vram at this point minimum. I would go for a desktop that can be upgraded much more easily. You could always buy a 10 GB vram card for now and then down the road consider getting something much better when they have even better ones.

@DailyClips009 5 месяцев назад

In model tab I see only one base model. how can I add more ?

@KLEEBZTECH 5 месяцев назад

How do you get more in the dropdown? If I understand then this would be the video is using locally. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-GgEEb2K0j7s.html

@diccionarioprimado 5 месяцев назад

Hello friend, can you make a video for the use of CPDS? Thanks in advance

@KLEEBZTECH 5 месяцев назад

I do cover CPDS in a couple of my other videos like the image prompt one. Not sure if you are looking for something more specific. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-2hW0M-Y1R6k.html

@diccionarioprimado 5 месяцев назад

@@KLEEBZTECH I saw it in your previous video, but it is not clear to me the difference in use between pyracanary and CPDS. There is not much documentation about it. Thanks in advance.

@KLEEBZTECH 5 месяцев назад

Honestly the difference in use is really dependent on the image and the results you want. They are technically for the same purpose to guide the structure but they just do it differently. I still have not determined 100% which is best for what.

@mick7727 5 месяцев назад

my outputs aren't saving suddenly. NVM it's because I changed the name of my focus folder. Simple things lmao.

@KLEEBZTECH 5 месяцев назад

lol. Always the simple things that get us.

@kenmasters734 5 месяцев назад

When i tried metadata in just had little error tabs everywhere

@KLEEBZTECH 5 месяцев назад

Any error messages in the CMD window?

@eugenmalatov5470 5 месяцев назад

if you have the same seed should the pictures with different colors not look the same?

@KLEEBZTECH 5 месяцев назад

Every tiny thing changes the ending image. So no it won't just change the color of the clothing. It would be great if that was the case.

@shubhambhanot985 Месяц назад

can you really help for how to use it fully cli based? please

@KLEEBZTECH Месяц назад

I don't use it that way so I could not really help much with that.

@punkouter23 5 месяцев назад

is there a way to make 20 picutures but change the value from one field? so I can gradually see how the value changing changes the pictuer. that make sense ?

@KLEEBZTECH 5 месяцев назад

Are you talking about the values for things like weight for image prompts or loras? Or do you mean the weights in the actual prompt? I don't know of any way for the first and with the new array feature you can't use parentheses inside the brackets or it makes the array feature not work.

@punkouter24 5 месяцев назад

@@KLEEBZTECH I tihnk the first... So the idea is I want to get 50 images with the weight slightly increasing from .5 to .8 for example.. that way I can understand better how it changes. .. so i guess no :(

@KLEEBZTECH 5 месяцев назад

Yeah not that I can think of. They did change it so you can now use prompt weights in the arrays but when talking about things like LoRAs or anything with a slider I have not found a way. If someone does have a way it would be nice. It might be possible if they add the ability to control that in the prompt but not sure if any plans for that.

@punkouter24 5 месяцев назад

that woudl be huge feature for me so I can set off 50 images and come back later.. and seems pretty easy to implement if I knew how to code@@KLEEBZTECH

@jeeadvanced4719 5 месяцев назад

can you please make a vidoe on how to increase fooocus generation speed, i have rtx4050 140watt 6gb vram and it still takes 30sec to generate

@KLEEBZTECH 5 месяцев назад

Is that using the quality setting? I know mine takes about 30sec total to generate and save the image with a 3070 8GB when generating on quality. I don't know if you are going to get it any faster unless you use less steps. 4050 with 6GB is not going to do very fast.

@CityUnderCeejDesigns 5 месяцев назад

Thanks for this! Quick Q: I launched mine and there was no update that happened. Advice / help?

@KLEEBZTECH 5 месяцев назад

Have you changed anything? Or using a different fork?

@CityUnderCeejDesigns 5 месяцев назад

@@KLEEBZTECH Nope. Just load it off my external drive. Not sure what what "fork" means. I get this when I launch every time: I:\Foocus\Fooocus_win64_2-1-831>.\python_embeded\python.exe -s Fooocus\entry_with_update.py Update failed. 'object not found - no match for id (dc5b5238c83c63b4d7814ba210da074ddc341213)' Update succeeded. [System ARGV] ['Fooocus\\entry_with_update.py'] Python 3.10.9 (tags/v3.10.9:1dd9be6, Dec 6 2022, 20:01:21) [MSC v.1934 64 bit (AMD64)] Fooocus version: 2.1.862 But it launches after 1 - 3 times of starting the program.

@KLEEBZTECH 5 месяцев назад

Have you tried downloading again to a different folder and trying? The nice thing is since it doesn't really "install" you can have multiple copies. I use them for testing myself. As for fork it just means other people have taken the code and made their own versions. I use the standard one myself 99% of the time but have tried some of the others.

@KLEEBZTECH 5 месяцев назад

I do keep copies of my config file backed up so I can use between them.

@sheldonventer2849 4 месяца назад

Im just wondering, how the fluff do you use embedding in Fooocus. I tried the method of (embedding:.....:1.1) but it does not actually seem to active.

@KLEEBZTECH 4 месяца назад

I have not used embeddings with Fooocus so not sure if it works like it should. And info seems very vague that I could find.

@sheldonventer2849 4 месяца назад

@@KLEEBZTECH Thank you for checking though, very cool :) But yeah. hope its something they work on as things develop

@KLEEBZTECH 4 месяца назад

@@sheldonventer2849if you figure it out and remember drop a comment here. I'm always trying to figure things out or find new ways of doing things.

@caseyj789456 5 месяцев назад

looking for 9:19 ! thanks

@dongytuetinh 5 месяцев назад

Do you have a colab version yet bro?

@KLEEBZTECH 5 месяцев назад

Not yet.

@majestic91 5 месяцев назад

Try to understand stable diffusion...so foocus is another software than sd? Is it local or is it google collab foocus? I really wanna learn all this but its soo much. Sd, lora, model, xl, 1.5, 2.1, checkpoints, weight, inpaint, faceswap, reactor, ..... how do u start?!

@KLEEBZTECH 5 месяцев назад

It is software and a user interface for using stable diffusion models. It can be done locally or on Google colab. I do have a bunch of videos covering all different aspects of it that you can check out. I have a video on installing locally but nothing on Google colab because I haven't messed around with that much.

@KLEEBZTECH 5 месяцев назад

Playlist ru-vid.com/group/PLSofK8pvSk5oHH9iNAAemQREADEY7CPp_

@bnthought5763 2 месяца назад

Sir ! When I'm running Fooocus. So I am writing last. press any key to continue. This problem is visible. Sir is there any solution? Please sir help with this.

@KLEEBZTECH 2 месяца назад

That really doesn't tell me much about your problem. If you are talking about message in CMD window there should be something before that which might indicate the issue.

@Eternal_23 5 месяцев назад

Guys, i have a question: after i've used Fooocus once and tried to restart same model it wouldn't start unless i've deleted downloaded checkpoint (aka model) and redownloaded it again, so how can i restart without redownloading the model?

@KLEEBZTECH 5 месяцев назад

Is it giving any error in the CMD window? It should only download if it doesn't see the model.

@Eternal_23 5 месяцев назад

@@KLEEBZTECH in case of a "run" bat file it starts quickly (model is in folder already) and once i try to generate something it's stuck at "Waiting for a task to start" in interface and "UNet ADM Dimension 2816" in CMD

@KLEEBZTECH 5 месяцев назад

Have you checked this? github.com/lllyasviel/Fooocus/blob/main/troubleshoot.md#adm-2816-then-stuck

@Eternal_23 5 месяцев назад

@@KLEEBZTECH now I did thanks to you, I'll make sure I have enough swap space, thanks 🤜🏻🤛🏻

@Eternal_23 5 месяцев назад

@@KLEEBZTECH Finally it works! Though since swap is on HDD it loads slowly, but that's ok, thank you so much bro!

@Alexdavidxs 5 месяцев назад

what is your pc configuration

@KLEEBZTECH 5 месяцев назад

I5 32GB Nvidia 3070 8gb vram

@AK47_64 5 месяцев назад

When will Stable Diffusion 3 be publicly launched, and will it be available on Fooocus?

@KLEEBZTECH 5 месяцев назад

I will check my crystal ball... 😉

@KLEEBZTECH 5 месяцев назад

Until they give more info it would just be a guess right now. I like your username by the way. Those are fun to shoot.

@AK47_64 5 месяцев назад

@@KLEEBZTECH 😎😇

@AK47_64 5 месяцев назад

@@KLEEBZTECH If stable diffusion 3 ever comes, will it appear in Fooocus!?

@KLEEBZTECH 5 месяцев назад

Probably.

@neonxp99 5 месяцев назад

They really, really, really have to add an erasure brush to the inpaint outpaint! It makes no sense having to step back 2,3,4,5, times or delete the whole mask, when you could just switch to the erasure bush.....and erase. Save a whole lota time and aggravation IMO.

@KLEEBZTECH 5 месяцев назад

Agree. 👍

@KLEEBZTECH 5 месяцев назад

I do think you are going to be seeing some changes to masking soon. Not sure if that will be part of it but I know it has been discussed from what I have seen on Github.

@b.s.design 5 месяцев назад

"Generate Image Grid for Each Batch" does not work (((((

@KLEEBZTECH 5 месяцев назад

I have never tried that before but I just tried it and doesn't seem to do what is expected...

@KLEEBZTECH 5 месяцев назад

Looks like you got that fixed...?

@b.s.design 5 месяцев назад

This button was responsible for making sure that in addition to the individual generated images, a common large file with all the images was also generated: drive.google.com/file/d/1uXzBVJzJO0y0Jyn8dzUoMnXDHCWDrBbD/view?usp=sharing@@KLEEBZTECH

@KLEEBZTECH 5 месяцев назад

Which is what I expected. Good job catching that bug and them fixing quickly. 👍

@b.s.design 5 месяцев назад

Yeah, the guys are great! Very promptly fixed everything!@@KLEEBZTECH