If you need the anime checkpoint/model you can just run the anime.bat file instead of the run.bat file. It will download the checkpoint automatically when you first run it. Otherwise you can check out this video on how to get and install them: kzitem.info/news/bejne/qJ17q5ZohGKbaqg
@xellostube
5 ай бұрын
12:53 I really really like that you left your trial and error in the video. I'll remember this better
@KLEEBZTECH
5 ай бұрын
Thanks. I find doing my videos this way gives people better idea of how things work.
@Radarhacke
2 ай бұрын
Wow! It's always amazing what's possible with fooocus. Thank you for today's revelation. ;-)
@techvishnuyt
5 ай бұрын
many videos are there about how to convert anime to real but none of them clearly covered real to anime, thank you for this!
@KLEEBZTECH
5 ай бұрын
Glad it was helpful!
@dirtydevotee
Ай бұрын
So, I'm looking at a ton of work here. Stable Diffusion is actually set up for this sort of thing. You want to use the token "(photograph:1.2)..." at the start of your prompt if you're anywhere other than Fooocus. You can also start with "cinematic (photograph:1.2)..." if you really want to land it in one gen, but you're decreasing the value of all your other tokens even further so it's a trade off. If you're using other style tokens, they go before "cinematic". So for instance, "tenebrism chiaroscuro cinematic (photograph:1.2) cat hunting mouse".
@vspolley
5 ай бұрын
Thanks, yet again!
@cmdr_talikarni
Ай бұрын
July 2024: I suggest using this with Pony v6 available in latest version 2.5.0+, you can use a prompt to convert without the extra steps (results can be hit or miss like usual).
5 ай бұрын
Thank you for the video! I hope the new SD 3.0 fixes this issue with hands when generating realistic images. I spend a lot of time fixing hands all the time.
@KLEEBZTECH
5 ай бұрын
I am not expecting it to be drastically better but hopefully some improvements.
@alexw9024
5 ай бұрын
Can we use sd3 in fooous?
@KLEEBZTECH
5 ай бұрын
Don't know if we will be able to. I would not expect it very soon though.
@AiPro84
5 ай бұрын
best explanation ever!Thanks
@KLEEBZTECH
5 ай бұрын
You are welcome!
@kmilo000
5 ай бұрын
Hi! thanks for the video, i was looking for how to convert anime into real, and this give me a lot of information about how to do it, and how to improve it! Long story short, i was in a doze, when I miss click the performance button on fooocus... I used the "Extreme Speed" option, and the results are AMAZING!, then you can use one of them that you liked and make the same process again but in "Quality" perfomance, :D hope this help to make another video so someone can make it better and better!
@MohamedElHanoun
4 ай бұрын
This Ai Is Awesome In A creepy Way Love U tuto keep Up ❤
@marcozisa6317
5 ай бұрын
Very informative. Thank you
@KLEEBZTECH
5 ай бұрын
Thanks for watching!
@stableArtAI
4 ай бұрын
we posted our video on this, it part of SD 1.5 transition between any checkpoints. Its best to use SDXL 1.0 with SD 1.5
@ilyanemihin6029
5 ай бұрын
Thanks! Very useful information! Focus is an amazing tool, you just need to know how to use it)
@KLEEBZTECH
5 ай бұрын
Very true!
@ilyanemihin6029
5 ай бұрын
@@KLEEBZTECH my dream is to use Fooocus via API, I'm using Fooocus from the RunDiffusion service now, but plan to install locally (8GB, Mac...) - to experiment with the API
@armauploads1034
Ай бұрын
Sir, it works, Sir! Thank you!
@KLEEBZTECH
Ай бұрын
You are very welcome!
@shamrockisland
5 ай бұрын
Not into anime but I like your presentation. Subscribed for more videos on the software thanks
@KLEEBZTECH
5 ай бұрын
Nor am I. lol. But I could not pass up the challenge of trying going back to realistic. It worked much better than I expected and also it can be used for other styles as well. Although some are more difficult than others. And thanks for the sub!
@b.s.design
5 ай бұрын
Thank You!
@KLEEBZTECH
5 ай бұрын
You're welcome!
@xellostube
5 ай бұрын
Try this totally different method that is usually used to upscale but I use it with more realistic checkpoints to convert anime/painting into photos. Automatic1111 --> Img2img and insert the image --> Control Net enable Tile/Blur and in another Control Net tab enable SoftEdge. Play with the resolution to increase it but keeping the propotions an dalso play with the Denoising Strenght I use Rev Animated as checkpoint
@KLEEBZTECH
5 ай бұрын
I will have to check it out next time I launch A1111. 👍Will admit I use it less and less these days.
@rbspider
3 ай бұрын
I didn't know they started out with real images. So that was nice to know. I thought it was all done with a description of the scene. Some technical questions. What are system requirements, can you run this on almost any pc? Is this done locally or on the Stable Diffusion website? I love the results.
@KLEEBZTECH
3 ай бұрын
You can run locally but it is suggested that you have a decent video card. If you try running on a PC without a dedicated GPU then you need tons of RAM (32GB I think) and it will be very slow. There is a recommended specs and an idea of how it will perform on the Github page. There is also the option to run online with services like Google Colab which I do have video on but the free tier does limit how much you can do with it each day. Just avoid websites that claim to be Fooocus and are really not. github.com/lllyasviel/Fooocus?tab=readme-ov-file#minimal-requirement
@KLEEBZTECH
3 ай бұрын
Also my video on using with Colab. Do note that there are currently issues with it sometimes not giving a public link but that seems to be getting better. kzitem.info/news/bejne/kW2b0qeCiodnm6Q
@rbspider
3 ай бұрын
@@KLEEBZTECH Thank you for taking the time to respond with all that detail.
@wernerblahota6055
5 ай бұрын
👍👍👍👍
@masterzed1
5 ай бұрын
Hello, I'm looking for an out painting that is similar to in Paint anything, meaning I want to create a mask within the drawing and then do the out painting. Can you provide an example using masks? I've tried doing this with in Paint anything, creating a mask over a person, and then modifying the background. However, it does it horribly, as it blends the original image as if it were part of the newly created image. I want to add a background with a robot behind a person, and when I input that description, what the AI does is only create a robot using the person as if it were part of a leg or a part of the robot.
@AsenPetrov
5 ай бұрын
Sorry for the question but Is there any way to set the exact number of Steps of Sampling? Thank you in advance. Great video!
@itsaboy213
5 ай бұрын
In the Advanced tab, Developer Debug Mode, after the sampler and scheduler settings, there is a Forced Overwrite of Sampling Step setting
@AsenPetrov
5 ай бұрын
@@itsaboy213 Thank you!
@cmdr_talikarni
Ай бұрын
Also the wording tends to weight heavily towards possible nsfw imagery. Having the same wording about her chest in multiple variations like he did at 13:00 also increases the weight of it as well, resulting in the image he got. I say this because when working with images of underage subjects (real or AI generated), this can result in undesirable "physical" effects that it is best to avoid (legally and morally).
@jeeadvanced4719
5 ай бұрын
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
@jeeadvanced4719
5 ай бұрын
It's loading in low vram mode altho I have 6gb vram
@masterzed1
5 ай бұрын
Definitely! I'm new to SD, but I might be able to help with increasing the cooling of your processor and video card, even doing a gentle overclock.
@masterzed1
5 ай бұрын
6:00 You have also a speed- button on the top right courner
@alee2077
4 ай бұрын
Hello, thank you so much for your video! says i have more than 100 anime style images would like to turn into realistic style, if there have any method to do batch generate in fooocus? or i can only do it 1 by 1??
@KLEEBZTECH
4 ай бұрын
No batch options for that in Fooocus.
@sickvr7680
5 ай бұрын
Hey! When convert from anime to photo u didnt uncheck LORA, its intended or an error? Thx! Very nice vid
@KLEEBZTECH
5 ай бұрын
sd_xl_offset is usually used with models like juggernaut so it was intentional. It is not necessary though.
@varunsharma2386
5 ай бұрын
hey can we use this in multiple photos at once ?
@KLEEBZTECH
5 ай бұрын
If you mean using different images as image prompts, you can but you will get a mixture of the images. You can get some interesting results.
@At0max
5 ай бұрын
and how to use this on A1111?
@KLEEBZTECH
5 ай бұрын
I am not familiar enough with A1111 to explain exactly how but A1111 should be able to with the img2img section. Some things I just prefer to use Fooocus for since it is easy to have them both.
@musicalstation101
5 ай бұрын
Please can do how to turn anime to drawing? Or how can i create drawing book using foocuse please 😢😢
@KLEEBZTECH
5 ай бұрын
You can do something similar and use the styles in Fooocus to try and get it to change to the style you like. I am not sure I fully understand what you mean by drawing book exactly.
@musicalstation101
5 ай бұрын
@@KLEEBZTECH I mean like foocuse give you drawing of eyes for exemple and give you shadow to draw the same thing
@KLEEBZTECH
5 ай бұрын
Maybe if you have some sort of example of a book that would help me understand fully. I think I have an idea what you mean but not 100% sure. @@musicalstation101
@musicalstation101
5 ай бұрын
@@KLEEBZTECH like drawing books for kids that sold online for exemple
@musicalstation101
5 ай бұрын
@@KLEEBZTECH and can you do how to turn like the pic u generated to hand drawing pic with fooocuse
@MrJanchoy1980
5 ай бұрын
@Kleebz Tech AI , many thanks , but how to install more Base Model , since i only have juggernautXL_v8Rundiffusion.safetensors
@KLEEBZTECH
5 ай бұрын
And here is the answer 👍: kzitem.info/news/bejne/qJ17q5ZohGKbaqg Will note that if you want to grab the anime checkpoint then you just need to run the anime.bat file and it will download it. Otherwise this video will explain how to get more.
Пікірлер: 68