bro was explaining every little thing like we are 5 year olds. i respect that a lot. not many creators realize that not all viewers are acquainted with even some of the basics
This is what I do with tutorials. I know MANY viewers do not need to hear it, but so many do! I want your grandma to be able to install after watching!
I got really annoyed when another RUclipsr tried to explain how to install an LLM on your home computer and he rushed through it and left out some key details. Ended up wasting over an hour of my time and I never did get it installed. 😢
Thank you so much for watching! PLEASE Check out our AI Community❤: discord.gg/bQgcbjs2Sg Links from today's video: ✩ AI Art Generation Software: github.com/lllyasviel/Fooocus ✩ 7-Zip: www.7-zip.org/download.html
How does the Fooocus fork compare to / differ from other forks such as Automatic1111, Invoke AI, ComfyUI, and Easy Diffusion? I know ComfyUI uses a node based UI but other than that how does it compare with the others including Fooocus? Is there a chart anyone has done giving the pros and cons of the popular forks like those?
"You need a GPU." ... Wrong!!! You need a powerful GPU that costs upwards of $500-$2000... Way to tell everyone about the minimum requirements by not telling anyone about the minimum requirements. 👌
I use MidJourney every single day and I went step by step with you to download this, I hope they stay on top of this and update it regularly. I can see this being my main go-to for AI Art Generation. Thanks Matt.
I know others have already said it but GREAT TUTORIAL! Not too fast, not too slow. Everything was explained in great detail with clearly spoken words. I thought you'd skip the 7 zip directions but you even explained that.
@@DiceDecides meh then that case ig my sagemaker method is bettter then 🥰. In 4 mins i can generate ateleast 20 pics without even installing anything haha
AI art seems to have undergone a quantum leap recently - as someone with degrees in art and computing I find it absolutely fascinating. You're my go to for info on it Matt. Your knowledge and enthusiasm are appreciated.
This is EXACTLY what I was looking for, thanks for making this tutorial! Most of the tutorials I've found go through all of these very basic steps like how to unzip a compressed file, and once it's installed, the final step is "now go do advanced AI programming stuff in ComfyUI with your computer engineering degree". Like nah bro, I have caveman brain, give me prompt box with button so can make pretty pictures.
Absolutely amazing! Thanks so much for this video. You explained how to install it so well that even I was able to get it all to work beautifully. I love all your RUclips videos, too!
I'm really curious to know if this image generator is better than the Midjourney one or if it's on the same level. 🤔 It's always great to see advancements in technology, and I appreciate you keeping us updated. 🙏 Keep up the good work! 👍
Matt, I copied the address of my desktop folder and then used the copy to here command and it finally worked. BUT, when the software finished running and copying all the files, instead of opening a window, it gave me a message saying that my NVIDIA board was TOO OLD, even though it has 8 gigs of RAM! This was on my laptop. I'm going to try it on my newer PC tower. GREAT VIDEO!
Thanks so much for letting me know about this Matt! I've been having fun and learning about loRAS and all that to get the kind of looks and styles I want it's amazing.
Hi Matt, I tested this software out today, and I must say, I am very impressed with the image generations it can produce. As you say in the video, it does stand up to or is better then Midjourney
Here's an interesting video idea for your future videos, Matt: Put all the best AI chatbots (Claude 2, Perplexity AI, ChatGPT, and Bing Chat) and compare all of them with different question prompts that will really challenge them, from creative prompts to productivity prompts, and at the end, only one AI chatbot remains victorious!
I've been playing around with this software since I got home. Got to say that this is incredibly impressive. First thing I tried was 'cat in tree' and while the second one had a really fat cat, the first one was really decent. Right now, I just tried 'Martial artist practicing on a mannequin'. While it ignored the mannequin, what came out looks like still frames right out of a martial artist movie. I think that they both have four fingers and a thumb on each hand. The ripples in their white outfits with black belt are incredibly detailed with just the right amount of blurring in the background to focus in on the guy. Just noticed his collar has dimples. Tried a sunset over the ocean and got some violent waters curving around some rocky terrain and it just looked amazing. Haven't seen much past some bow legging going on with a guy playing an arcade game. The engine even varied the games on each screen and did its best to faux name them. Why, this looks even better than some of those on-line engines that produce far more oddities. Makes me wonder if in a year or two, could I just plug in a microphone and ask these people questions... and not only get full motion, but full historical backgrounds? I just can't believe this is totally free. It's just too good.
I downloaded this today. This is absolutely amazing. You definitely need a lot of memory. I have 32gb RAM and a Nvidia GTX 3060. 12GB video card. It comes close to maxing out my PC at times. I love the quality of the images and the fact that it is free. I use Adobe Photoshop and now that Firefly is out of Beta, they are already talking about limiting the number of generations each month.
Matt! I know you get this a lot but THANK UUUUUUUU I LOVE U :3 It didn't work at first cause it said my GPU was too old but then I realized I just didn't download the latest driver lol THANK YOUUUUUUUUU
Pro Tip: A shortcut to the task manager is pressing CTRL Shift and Esc. Also while you have it open start 7zip then go to details tab, select 7gz from the list, right click on it and set priority to high. This greatly speeds it up and shouldn't cause too much lag. I don't advise real time but it will be slightly faster than high.
Great video. Love how simply you laid everything out. Question, when copying the files over with 7zip the process gets hung at 31%. I have left it running for 5 hours, restarted, deleted everything and re-downloaded and restarted again but it hangs at 31% every time. Running 4060ti with 8gb gpu and 32gb ram. Do you have any suggestions?
tried this on my lower end browsing PC - just a 4gb gtx1080ti - runs slow but still great results - takes maybe 20 minutes to generate, but when the results are so reliably good quality, it doesn't even matter. Thanks for the video
Im getting an error when i run the program. [WinError 10054] an existing connection was forcibly closed by the remote host and press an key to continue which closes the window. also shows an error in line 557 and 749 (in http_error_302). any tips on how to resolve this? tia!
This Fooocus program works. I have tried all Stable diffusion with auto1111 with civit ai models..somewhat I cannot get a good result. However, this simple bat file works great and the outcome is amazing. Just like MJ. and I tested text generating and it generate text clearly too. Thank you Matt!!!
There was an update regarding AMD cards. I tried to folow the guidelines. Had to edit the run.bat file adding some lines. Unfortunately it only managed once to open the program and got stuck there. Next attempts where futile. Something was out of date in the file or didn't correspond. I am an amateur user not a programer, so I didn't make much sense of it. So I eneded up deleting the whole file since I couldn't make it work. Maybe in the future they upload a working update for AMD users.
This is so cool except I have two GPU's (Intel and Radeon) and at this point the Fooocus software will not run with those GPU's. Do you think that they may include them at a time or is this so CUDA centric that it would mast likely no happen?
Just a quick question, how would I uninstall this if I no longer wanted it on my pc? Asking BEFORE I install as I have been caught out before with no way to remove software
Thanks man, My right hand is fucked up so its hard to type but I wanted to say thankyou. Ive been waiting for something to drop my midjourney membership!
when im coying the files from 7-zip to my new folders my download is stuck at 31% after 2-3 seccond of copying and nothing happen even after hours, idk whats the problem.. (i dont have a low spec pc btw i got hi-end monster so im confuse) ..
Hi Matt, this is great! Thank you for sharing your knowledge. Do you have any idea when this is going to be available for Mac? Or any idea how to do this on Mac? Thank you again so much.
Awesome tutorial! Any tips on accessing the web UI from a different computer on the same network? I want to host it on my computer but let my kids have access for story book creation from their laptops.
Please please make a few videos covering the advanced features of this software for us n00bs. We have to start somewhere, and this looks like the right software and the right creator. You vids are top tier, thank you!
Nicee! Thanks for do this Matt. I downloaded the previous version of this software, If i download the new 2.0 version it download all the components again or only upgrade the previous one in my computer?
how fast is the computer youre on in this video? how long did it take to generate an hd pic? i have such a paltry graphics card, only 4gb vram. would it take an hour to do each generation in this?
You should make a turorial on how to add own styles, choose the folder where it puts the images and set a custom models folder in case you want to share models with other apps. You can also set a custom startup config.
How do you get it to use the Input facial Image in it's AI generation. I ticked the check box, dragged a face imagine on it, and started creating but it seems to completely ignore the input image. Do I have to put anything inside the image prompt to get it to use the input image? Also, how do you add more than one input image... like a whole model of facials images like Leonardo is able to do? The images genterated are interesting, but none of them are usable because theya ren't based on my Input image at all... (?) There is one nice thing about it... it gives my 3070 GPU which sits idle 99.9% of the time something to grind it's teeth on in the background while I'm doing other stuff on the comptuer without impacting my computers perforanace at all. It's like having a second computer working inside your computer...
Happen to know off hand how to get it to dump all info in a drive other can c:\? I tried a local one a while back, and after generating 2 images, it had filled my c drive with data. I'd like to mess around with some of these, but I need to be able to keep all the data contained somewhere other than my main drive.
I have 7800xt i get the part where it is supposed to be slower than nvidia but why it's crashing 😑😐😑😐😑 Windows is working once in low resolution (only one image) than restart Arch Linux segment core dump I make a temporary swap on pendrive
How does this compare with ComfyUI running sdxl? I guess it's the same thing but here you have less control since you don't have nodes to create complex workflows but at the same time they already have included upscaling, in-panint and maybe other stuff that I didn't catch without having you to create these processes with nodes.
it's not working. after downloading. the user interface, when prompting and hitting generate, it keeps loading, but nothing. in the command I did see that the second download stopped automatically at like 35%. but when I run it again, it won't load the rest. aby Idea what went wrong and how to solve it?
I have Installed on PC with AMD GPU, slow but it works! Follow Matt's instructions and then edit run.bat file to .\python_embeded\python.exe -m pip uninstall torch torchvision torchaudio torchtext functorch xformers -y .\python_embeded\python.exe -m pip install torch-directml .\python_embeded\python.exe -s Fooocus\entry_with_update.py --directml pause
The run-with box never popped up. But the 7-Zip did show up on the right click. I had to experiment with the options but I waited until I got the files you got. BTW, that old school Bat. field did not show but the slightly newer school run. file did show. The program isnt fast but it works albeit a little slow with my 8 gig video card. Ah well, it works and its free for now and the images are spectacular.
bro was explaining every little thing like we are 5 year olds. i respect that a lot. not many creators realize that not all viewers are acquainted with even some of the basics
That’s why EILI5 was created. Someone somewhere needs basic info, depending on the subject matter.
This is what I do with tutorials. I know MANY viewers do not need to hear it, but so many do! I want your grandma to be able to install after watching!
ay man the first thing I would've done is closed that black window like WHY ISN'T IT WORKING LOLOL
I got really annoyed when another RUclipsr tried to explain how to install an LLM on your home computer and he rushed through it and left out some key details. Ended up wasting over an hour of my time and I never did get it installed. 😢
@@AmazingArends people do not realize the importance of explaining!
Thank you so much for watching! PLEASE Check out our AI Community❤: discord.gg/bQgcbjs2Sg
Links from today's video:
✩ AI Art Generation Software: github.com/lllyasviel/Fooocus
✩ 7-Zip: www.7-zip.org/download.html
👋
I have shared this video on both the artificial intelligence group and the AI art group on Gab! Very cool! 🎉
How does the Fooocus fork compare to / differ from other forks such as Automatic1111, Invoke AI, ComfyUI, and Easy Diffusion? I know ComfyUI uses a node based UI but other than that how does it compare with the others including Fooocus?
Is there a chart anyone has done giving the pros and cons of the popular forks like those?
NO LEMON GUY!!?? what is this??? I was hoping you make the lemon character on the beach wearing sun glasses! 😉
"You need a GPU." ... Wrong!!! You need a powerful GPU that costs upwards of $500-$2000...
Way to tell everyone about the minimum requirements by not telling anyone about the minimum requirements. 👌
I use MidJourney every single day and I went step by step with you to download this, I hope they stay on top of this and update it regularly. I can see this being my main go-to for AI Art Generation. Thanks Matt.
I know others have already said it but GREAT TUTORIAL! Not too fast, not too slow. Everything was explained in great detail with clearly spoken words. I thought you'd skip the 7 zip directions but you even explained that.
No skimping on these tutorials!
not having to worry about using up credits is huge, thanks for making a thorough tutorial on this, I'll have to install it
does it need very high system specs?
@@ayoww1000 not at all, my pc is over a decade old and well... it does take like 4 mins to generate an image but it works!
@@DiceDecides meh then that case ig my sagemaker method is bettter then 🥰. In 4 mins i can generate ateleast 20 pics without even installing anything haha
Thanks for the tutorial. This is a beautiful art generator, best open source I've seen so far.
Happy you like it! Always ready to support the community ❤
Loved the "new" style of video, Matt! You got some good teaching skills.
Keep it up !! 🇧🇷❤
Thank you so much!
Whats new?
Totally unacceptable that you don't have more subscribers! Thanks man, top info as usual :)
Thank you! Greatly appreciate the kind words!
AI art seems to have undergone a quantum leap recently - as someone with degrees in art and computing I find it absolutely fascinating. You're my go to for info on it Matt. Your knowledge and enthusiasm are appreciated.
This is EXACTLY what I was looking for, thanks for making this tutorial! Most of the tutorials I've found go through all of these very basic steps like how to unzip a compressed file, and once it's installed, the final step is "now go do advanced AI programming stuff in ComfyUI with your computer engineering degree". Like nah bro, I have caveman brain, give me prompt box with button so can make pretty pictures.
You’ve given me exactly what I was looking for and desired, plus an extra bonus of being free!
This video is exactly why I'm subbed to you, very helpful and awesome stuff presented to me as it comes out, great stuff Matt cheers
Thanks for the sub!!!
Almost every text2img out there beats MJ in controlling the image you want to create 😊
Great vid Matt. Thanks
So true!
I don't comment often but i had to say that the video was concise and useful. appreciate it.
Absolutely amazing! Thanks so much for this video. You explained how to install it so well that even I was able to get it all to work beautifully. I love all your RUclips videos, too!
Thank you so much that means a lot!
Bro this video was amazing, very well detailed, love it.❤
Appreciate it!
I'm really curious to know if this image generator is better than the Midjourney one or if it's on the same level. 🤔 It's always great to see advancements in technology, and I appreciate you keeping us updated. 🙏 Keep up the good work! 👍
Matt, I copied the address of my desktop folder and then used the copy to here command and it finally worked. BUT, when the software finished running and copying all the files, instead of opening a window, it gave me a message saying that my NVIDIA board was TOO OLD, even though it has 8 gigs of RAM! This was on my laptop. I'm going to try it on my newer PC tower. GREAT VIDEO!
Thanks for the information! Maybe update drivers or even the bios?
Works! Here on November 7th 2024 on Windows 11 this worked fantastically with no issues, thank you so much man. :]
This is fantastic. Hope to see this program further developed!
Hmmm it tends to always stay stuck on "initialize" and it's very very slow
WOW!
Big Fanx for introducing this and making things so easy to handel !
Of course!
Thanks so much for letting me know about this Matt! I've been having fun and learning about loRAS and all that to get the kind of looks and styles I want it's amazing.
If it's compatible with Mac, it can also be extended to run on other Apple devices such as the iPad with M1 or M2 chips, through mobile porting
Hi Matt, I tested this software out today, and I must say, I am very impressed with the image generations it can produce. As you say in the video, it does stand up to or is better then Midjourney
Great to hear!
One of the best videos you've ever made. Thanks, man.
Thank you! I worked hard on this tutorial to make sure it’s as easy as possible
Fooocus also works just fine on macOS following the install instructions given for Linux.
Really? That’s surprising
Yesssss! So I have to try cause I'm on Mac too ... Mac Osx is unix based by the way
What I find the most amazing is that a software that can generate an image of anything under the sun occupies less than 7 GB.
Insanity
Here's an interesting video idea for your future videos, Matt: Put all the best AI chatbots (Claude 2, Perplexity AI, ChatGPT, and Bing Chat) and compare all of them with different question prompts that will really challenge them, from creative prompts to productivity prompts, and at the end, only one AI chatbot remains victorious!
You could totally do this with nat.dev!
@@MattVidProjust tried it. It's pay to play now
Such a good tutorial that I only needed to watch half of it, I know that sounds weird but it was really simple, thank you! -DMP
Been using this for a while now. Really good program
After downloading Press any key to continue shows. The pop up menu does not appear.
Thanks for another great video Matt. I'm setting it up right now. The MattVidPro channel 🤘🤘🤘
Hope you enjoy it! Thank you!
ONLY thing i dislike about this video the most... i cant like this more than once!!!! i am so happy about this
So happy to help!
I've been playing around with this software since I got home. Got to say that this is incredibly impressive. First thing I tried was 'cat in tree' and while the second one had a really fat cat, the first one was really decent. Right now, I just tried 'Martial artist practicing on a mannequin'. While it ignored the mannequin, what came out looks like still frames right out of a martial artist movie. I think that they both have four fingers and a thumb on each hand. The ripples in their white outfits with black belt are incredibly detailed with just the right amount of blurring in the background to focus in on the guy. Just noticed his collar has dimples. Tried a sunset over the ocean and got some violent waters curving around some rocky terrain and it just looked amazing. Haven't seen much past some bow legging going on with a guy playing an arcade game. The engine even varied the games on each screen and did its best to faux name them. Why, this looks even better than some of those on-line engines that produce far more oddities. Makes me wonder if in a year or two, could I just plug in a microphone and ask these people questions... and not only get full motion, but full historical backgrounds? I just can't believe this is totally free. It's just too good.
Is there any restricted words?
It feels like the early days of the internet.
We are in the early days of the internet lol
This video makes me subscribe!
Thank you so much!
Ngl, I like his ai text to speech tone of voice. 10/10 🏆
What camera do you use for your videos?
Canon EOS RP w 1.8 50mm
Thanks bro! you're a life saver, I dont really have alot of time to spend on my PC cause of college, so this really saved me alotta time :)
I downloaded this today. This is absolutely amazing. You definitely need a lot of memory. I have 32gb RAM and a Nvidia GTX 3060. 12GB video card.
It comes close to maxing out my PC at times. I love the quality of the images and the fact that it is free. I use Adobe Photoshop and now that Firefly is out of Beta, they are already talking about limiting the number of generations each month.
I sure hope they don't do that!
Matt! I know you get this a lot but THANK UUUUUUUU I LOVE U :3 It didn't work at first cause it said my GPU was too old but then I realized I just didn't download the latest driver lol THANK YOUUUUUUUUU
Pro Tip: A shortcut to the task manager is pressing CTRL Shift and Esc. Also while you have it open start 7zip then go to details tab, select 7gz from the list, right click on it and set priority to high. This greatly speeds it up and shouldn't cause too much lag. I don't advise real time but it will be slightly faster than high.
Great video. Love how simply you laid everything out. Question, when copying the files over with 7zip the process gets hung at 31%. I have left it running for 5 hours, restarted, deleted everything and re-downloaded and restarted again but it hangs at 31% every time. Running 4060ti with 8gb gpu and 32gb ram. Do you have any suggestions?
tried this on my lower end browsing PC - just a 4gb gtx1080ti - runs slow but still great results - takes maybe 20 minutes to generate, but when the results are so reliably good quality, it doesn't even matter. Thanks for the video
1080 Ti has 11 gb of ram
Im getting an error when i run the program. [WinError 10054] an existing connection was forcibly closed by the remote host and press an key to continue which closes the window. also shows an error in line 557 and 749 (in http_error_302). any tips on how to resolve this? tia!
This Fooocus program works. I have tried all Stable diffusion with auto1111 with civit ai models..somewhat I cannot get a good result. However, this simple bat file works great and the outcome is amazing. Just like MJ. and I tested text generating and it generate text clearly too. Thank you Matt!!!
Thanks man! Outstanding channel. Subscribed.
There was an update regarding AMD cards. I tried to folow the guidelines. Had to edit the run.bat file adding some lines. Unfortunately it only managed once to open the program and got stuck there. Next attempts where futile. Something was out of date in the file or didn't correspond. I am an amateur user not a programer, so I didn't make much sense of it. So I eneded up deleting the whole file since I couldn't make it work. Maybe in the future they upload a working update for AMD users.
Is it uncensored so one could generate images of soldiers with rifles for instance?
This is so cool except I have two GPU's (Intel and Radeon) and at this point the Fooocus software will not run with those GPU's. Do you think that they may include them at a time or is this so CUDA centric that it would mast likely no happen?
You are the best! Using it now! Thank you Matt!
Amazing description and channel overall. I have an AMD GPU (7090xtx) and it works for anyone that was wondering. 👍
good thing i watched that video ! thank you matt !
Glad you liked it!
Just a quick question, how would I uninstall this if I no longer wanted it on my pc? Asking BEFORE I install as I have been caught out before with no way to remove software
Thanks man, My right hand is fucked up so its hard to type but I wanted to say thankyou. Ive been waiting for something to drop my midjourney membership!
happy to help! hope your hand feels better!
This earned you a sub buddy :)
U just changed my entire AI art experiance!!!!!
Yo Fkn had me at hello buddy. This is amazing! Tytytyvm!
Me watching the full tutorial knowing my PC doesn't run it: :,D
Thanks dear , but I always get an error: mat1 and mat2 must have the same dtype ???? any help
Same, it's probably some kind of a bug but tis very annoying
Yes after the setp 20 from tatal 30 steps ,, it stop and create that error !!!!
@@soulstream1408
Same,been tryingbut havent figured out what to do about it yet, if u find out pls let me know
when im coying the files from 7-zip to my new folders my download is stuck at 31% after 2-3 seccond of copying and nothing happen even after hours, idk whats the problem.. (i dont have a low spec pc btw i got hi-end monster so im confuse) ..
Same for me ever figure out the issue?
Nice work, as usual. Thanks.
Thank you for watching!
You've given me the best gift of 2023
this is insane, like giving the city's keys level!
Hi Matt, this is great! Thank you for sharing your knowledge. Do you have any idea when this is going to be available for Mac?
Or any idea how to do this on Mac? Thank you again so much.
Awesome tutorial! Any tips on accessing the web UI from a different computer on the same network? I want to host it on my computer but let my kids have access for story book creation from their laptops.
you can also press CTRL + SHIFT + ESC to quickly get the task manager
I dont get that downloading bar after opening run dot bat. It says its downloading but the image generator white window never pops up
THIS IS HOW EVERY TUTPRIAL VIDEO SHOULD BE LIKE
Please please make a few videos covering the advanced features of this software for us n00bs. We have to start somewhere, and this looks like the right software and the right creator. You vids are top tier, thank you!
Is this model uncensored?
will this take advantage of the new NPU,s like the new amd AI cpus and snapdragon ai npu?
Nicee! Thanks for do this Matt. I downloaded the previous version of this software, If i download the new 2.0 version it download all the components again or only upgrade the previous one in my computer?
how fast is the computer youre on in this video? how long did it take to generate an hd pic? i have such a paltry graphics card, only 4gb vram. would it take an hour to do each generation in this?
Apparently support for AMD GPUs was already added on Linux, for anyone wanting to try it you could use virtual box
Generation will be very slow on 4GB GPU. You should to have at least 8GB for good enough performance.
I have an 8gb 3070 and its barely 15% done a single image after 5 minutes
can you tell me the commands i should put in the prompt before i go for a design ?
You saved my day. I get the same results faster than I used with midjourney and playground. Do we have a similar alternate for elevan labs audio?
You should make a turorial on how to add own styles, choose the folder where it puts the images and set a custom models folder in case you want to share models with other apps. You can also set a custom startup config.
You saved me some cash, thank you so much! Got a new subscriber
How do you get it to use the Input facial Image in it's AI generation.
I ticked the check box, dragged a face imagine on it, and started creating but it seems to completely ignore the input image.
Do I have to put anything inside the image prompt to get it to use the input image?
Also, how do you add more than one input image... like a whole model of facials images like Leonardo is able to do?
The images genterated are interesting, but none of them are usable because theya ren't based on my Input image at all... (?)
There is one nice thing about it... it gives my 3070 GPU which sits idle 99.9% of the time something to grind it's teeth on in the background while I'm doing other stuff on the comptuer without impacting my computers perforanace at all. It's like having a second computer working inside your computer...
So cool!!Thanks for sharing.
Thanks for watching!
Happen to know off hand how to get it to dump all info in a drive other can c:\? I tried a local one a while back, and after generating 2 images, it had filled my c drive with data. I'd like to mess around with some of these, but I need to be able to keep all the data contained somewhere other than my main drive.
I've just been testing this. It's really good. Thanks for sharing @MattVidPro
Glad you like it!
I have 7800xt i get the part where it is supposed to be slower than nvidia but why it's crashing 😑😐😑😐😑
Windows is working once in low resolution (only one image) than restart
Arch Linux segment core dump I make a temporary swap on pendrive
I normally train lora for civitai can you train loras on this too. Like I want to make a checkpoint.
Brother is this still work like this and offline?
Any better offline models or updated ones i can install locally and use freely bro
Unfortunately, the photos are not generating :(
Cool deal. Would love to add more models. The drop down seems to indicate I can but no instructions on doing so.
Great Post - thanks a lot. What is the Browser you need for this?
Error checking version for opencv-contrib-python-headless: No package metadata was found for opencv-contrib-python-headless
Thank you for a great tutorial! How do you run this without re-downloading the model each time? Thanks! 🙂
Brilliant presentation! One thing, how will I update this in the future? Is there a set way? It’s one thing that’s worried me.. Any idea? Cheers!
How does this compare with ComfyUI running sdxl?
I guess it's the same thing but here you have less control since you don't have nodes to create complex workflows but at the same time they already have included upscaling, in-panint and maybe other stuff that I didn't catch without having you to create these processes with nodes.
This is much easier to install and use
it's not working. after downloading. the user interface, when prompting and hitting generate, it keeps loading, but nothing. in the command I did see that the second download stopped automatically at like 35%. but when I run it again, it won't load the rest. aby Idea what went wrong and how to solve it?
I would try deleting everything and redownloading
Bad GPU or not enough RAM can be a reason. Try to generate smaller images.
My default unzipper program didn't unzip it properly / completely and that caused it not to run right the first time I tried it
I have Installed on PC with AMD GPU, slow but it works! Follow Matt's instructions and then edit run.bat file to
.\python_embeded\python.exe -m pip uninstall torch torchvision torchaudio torchtext functorch xformers -y
.\python_embeded\python.exe -m pip install torch-directml
.\python_embeded\python.exe -s Fooocus\entry_with_update.py --directml
pause
Amazing! Thanks, Matt!
Thanks!
The run-with box never popped up. But the 7-Zip did show up on the right click. I had to experiment with the options but I waited until I got the files you got. BTW, that old school Bat. field did not show but the slightly newer school run. file did show. The program isnt fast but it works albeit a little slow with my 8 gig video card. Ah well, it works and its free for now and the images are spectacular.