Installing Stable Diffusion used to be a pain involving downloading a bunch of different software, packages, and python - Today it's literally almost a single step-!! For those of you interested in learning how to train your own AI model with Stable Diffusion FAST, I have a playlist for it right here: ruclips.net/video/2Nvpbvnj_u0/видео.html If you're having trouble, we might be able to help you on Discord, you can find our channel here: discord.com/channels/1138979360943120445/1204654189897322557 Here's the link to the GitHub: github.com/AUTOMATIC1111/stable-diffusion-webui
Hi @TheRoyalskies - I got this message at the end of running the software - Traceback (most recent call last): File "C:\AI\webui\launch.py", line 48, in main() File "C:\AI\webui\launch.py", line 39, in main prepare_environment() File "C:\AI\webui\modules\launch_utils.py", line 384, in prepare_environment raise RuntimeError( RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check, what do I do?
@@NaviaDaddyIssuee Yes, as long as the final image generated is original and unique you should be fine. But, If you generate an image of something popular and famous though like Pikachu, then of course you can't sell that legally. You can use AI to create your own "Electric Rat" with its own unique design if you want, but yeah, make sure the design is fresh and original.
I search for this type video. Fast and straight forward. Sick n tired of people talking too much thinking they can make tons of money with long video. RUclips algorithm is changing. Thanks bro
My freaking God. I literally spent 2 hours fighting around errors and dozens of long tutorials to get it work only to find out that this 40 sec video is the solution. THANK YOU
Not sure if you still read newer comments but thank you so much for this! I watched three other tutorials that failed but this worked flawlessly! Thank you!
Yeah, this could do with an installer - we are artists* at the end of the day! *Well, its more like being soeone who commissions artwork and the tireless AI creates generally what you want
@@alihp7844 go to the system file in your installed folder , then enter the python file(D:\AI\system\python) ,write cmd in the location bar (not the search bar but the one where it says the location youre in),press enter and write this command python -m pip install setuptools==69.5.1
omg THANK YOU so much! I am so bummed with Playground AI eliminating their image generator, so needed an alternative. This helped me get up and running with stable diffusion quickly and easily!
I keep clicking "run" it opens the window with the black background, then it does it's thing, finally it says "press any key to continue" when i press any key, it closes -- stable diffusion is nowhere to be found I need help :(
I got this error can you please help? Python 3.10.6 (tags/v3.10.6:9c7b4bd, Aug 1 2022, 21:53:49) [MSC v.1932 64 bit (AMD64)] Version: v1.8.0 Commit hash: bef51aed032c0aaa5cfd80445bc4cf0d85b408b5 Traceback (most recent call last): File "C:\AI\sd.webui\webui\launch.py", line 48, in main() File "C:\AI\sd.webui\webui\launch.py", line 39, in main prepare_environment() File "C:\AI\sd.webui\webui\modules\launch_utils.py", line 386, in prepare_environment raise RuntimeError( RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check Press any key to continue . . .
I'm not incompetent to learn doing stuff the hard way when it comes to things that involve command prompt... but... boy was I really procrastinating on SD. Thanks for helping show that I didn't have to hack into the mainframe to get this onto my PC. xD
Subscribed. Mans went straight to the point and didn't try bullshitting his audience or spewing out the word "Free" when there is token systems and paywalls behind certain services.
I got this error RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check Any way to fix this?
@@aberryth_dev I don't know you and I have no idea what's going in your life, but you matter bro please never forget this.I really mean it.Thank you so much for this detailed and easy to understand explanation, it helped me and now it works.I wish you all the best in life.
@@aberryth_dev Any other ideas? GPU driver up to date, I also tried to no avail with Python 3.9. The earlier more longer process gave me the same problem, I was able to get as far as this error when I tried to import.torch: A module that was compiled using NumPy 1.x cannot be run in NumPy 2.1.0 as it may crash. To support both 1.x and 2.x versions of NumPy, modules must be compiled with NumPy 2.0. Some module may need to rebuild instead e.g. with 'pybind11>=2.12' I though Python 3.10.6 used the Numpy module version 1.x. ? I have no other versions of Python on this machine. Is the Geforce GT 430 just cursed?
@@rainbowrae267 it looks like the current version of numpy is too recent for the other dependencies. First, try running update.bat (obvious, I know, but I’ve got to check) Using the same method as above, try -reinstall-numpy==1.26.4 Or if that doesn’t work: -reinstall-numpy-upgrade Or just -reinstall-numpy I’ve never seen that before though, so honestly I don’t know
@@rainbowrae267 When I had that problem i had moved everything into the "Videos" folder on my computer. It was suggested in a reddit thread I found and it worked for whatever reason. So try moving your "stable-diffusion-webui" folder, Python app and "Git-2.43.0-64-bit" app into your "Videos" folder. When I did that it worked
I did everything, once I clicked on "run" the black screen opened and it did it's stuff, then it said "press any key to continue" I pressed and the window closes. How do i run stable diffusion?
i keep get that error : ImportError: cannot import name 'packaging' from 'pkg_resources' (D:\sd.webui\system\python\lib\site-packages\pkg_resources\__init__.py) any help please?
I had the same problem, it occurs when setuptools version is 70.0.0. You must run the cmd on the system folder of webui and enter "python -m pip install setuptools==69.5.1".
the best you could do for that (i think) is make a background. first then img2img that use uh i forgot the name of the tool. its basically like a cutout the place you want the girl to be. then itll fill that
hey Royal Skies, thanks for the help. I tried both tutorials and none of them worked. The installion is sucessfull but the stable diffusion simply doesn't work. it gives all sort of errors when i try to generate anything, like this OutOfMemoryError: CUDA out of memory. 😔
same here... i used to have beautifulll stabe diffusions like a year ago, without any VAE etc i was just using anything v3 and it was beautifull but now the same checkpoint etc the same and everything is ugly... did you find solution?
Means that you might have to update python. or you need a lower python version. Just try to update python first if didn't work then try to install a lower version
It doesn't work. It says: "ERROR: Could not install packages due to an OSError: [WinError 5] Access is denied Consider using the ' -- user' option or check the permissions."
u made it look so easy but its taking me a lot longer because it took me 15minutes to figure out that i need to install a third party happ called git so it can use the git command, and then when u launch it it downloads dependancies like pytorch which in total is 2.5 gb and thats another 15minutes on my internet speed, also i had to run them as admin to get them to do anything so just over 30 minutes in total at this rate
yes and no. You sill will have a white (or w/e color border) around the hexagonal shape generation. The final image will have 4 sides, with a hexagonal shape interior generated image. There are different ways to do this. One way in inpainting.
@@highcommander2007 As you mentioned regarding inpainting. Can you give an example prompt to create image on a hexagon shaped canvas . Your reply will be greatly appreciated. Thanks in advance.
I'm getting this error, when trying to do it. FileNotFoundError: No checkpoints found. When searching for checkpoints, looked at: - file C:\Ai Art\sd.webui\webui\model.ckpt - directory C:\Ai Art\sd.webui\webui\models\Stable-diffusionCan't run without a checkpoint. Find and place a .ckpt or .safetensors file into any of those locations. Time taken: 0.1 sec.
hello i keep getting this error. i followed your tutorial exactly. how do i fix this. NansException: A tensor with all NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check.
Just ask AI my brutha.. :D The error message you provided appears to be related to a software exception in a program or library called Unet. Unet is commonly used in the context of deep learning, particularly in image segmentation tasks. The error message indicates that a tensor (a multi-dimensional array) containing NaNs (Not-a-Number) values was produced during the execution of the Unet model. NaNs typically occur as a result of undefined mathematical operations, such as division by zero or taking the square root of a negative number. The message suggests a few potential solutions: Set "Upcast cross attention layer to float32" option: If there is an option in the settings of Unet related to upcasting cross attention layers to float32 (single precision floating point numbers), enabling this option might resolve the issue. Use the --no-half command-line argument: This command-line argument suggests disabling the use of half-precision floating-point numbers (float16), which may not be supported by your hardware. Use the --disable-nan-check command-line argument: This command-line argument suggests disabling the check for NaN values altogether. However, this might not be a recommended solution as it could potentially mask underlying issues in the model or data. Consider trying these suggested solutions based on your specific use case and requirements. If you're unsure about how to implement these solutions or if the issue persists, you might need to seek further assistance from the developers of Unet or from a community/forum where Unet users discuss and troubleshoot similar issues.
⭐⭐⭐⭐⭐ This guy needs to do a Ted Talk on RUclips video tutorial making. Didn't even waste a second to ask the viewer to subscribe or like. I subscribed to the channel and liked the video simply because of that fact🤪
A tensor with all NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check. anyone know how to fix this
If you are having trouble installing Stable Diffusion, try stopping Windows updates in the background in the task manager. Once you have done that, try running the installation again.
Hello RUclips reader. I following the video and got the following message "RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check Press any key to continue . . ." I have a Intel UHD Graphics 630 Please Help.
I keep clicking "run" it opens the window with the black background, then it does it's thing, finally it says "press any key to continue" when i press any key, it closes -- stable diffusion is nowhere to be found I need help :( Did you ever fix it???
the astronomical number of star on automatic1111 that you can look up on google is a nice hint. 128k star as of today. if that can't be trusted on the official github website i don't know what is.
![Stable Diffusion Tutorial - How to Install, How to Use [Nvidia & AMD]](http://i.ytimg.com/vi/i0_QKju-3PQ/mqdefault.jpg)
Installing Stable Diffusion used to be a pain involving downloading a bunch of different software, packages, and python - Today it's literally almost a single step-!!
For those of you interested in learning how to train your own AI model with Stable Diffusion FAST, I have a playlist for it right here:
ruclips.net/video/2Nvpbvnj_u0/видео.html
If you're having trouble, we might be able to help you on Discord, you can find our channel here:
discord.com/channels/1138979360943120445/1204654189897322557
Here's the link to the GitHub:
github.com/AUTOMATIC1111/stable-diffusion-webui
do you plan on showing how to use the program? I'm don't know how to make good image.
One question can you use the pictures without any harm to law like copyright strike and so on are they free to use for business like print it on shirt
Hi @TheRoyalskies - I got this message at the end of running the software - Traceback (most recent call last):
File "C:\AI\webui\launch.py", line 48, in
main()
File "C:\AI\webui\launch.py", line 39, in main
prepare_environment()
File "C:\AI\webui\modules\launch_utils.py", line 384, in prepare_environment
raise RuntimeError(
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check, what do I do?
Can I use it for commercial use
@@NaviaDaddyIssuee Yes, as long as the final image generated is original and unique you should be fine. But, If you generate an image of something popular and famous though like Pikachu, then of course you can't sell that legally. You can use AI to create your own "Electric Rat" with its own unique design if you want, but yeah, make sure the design is fresh and original.
Wow..a tutorial without having to show how to connect a mouse, open a web page, or first check if the PC has a GPU. I am grateful, my esteemed sir.^^
You've got to mine the silicon for making your transistors first!
@@LagMasterSam He didn't show how to build my own excavator and my own smelting facility to process the silicon. 0/10 tutorial
@@TheNerogarden Nor did he show us how to prospect for silicon deposits to begin with. I'll be unsubscribing.
In a world... full of algorithmically-optimized 10 minute long tutorials on how to click a link... ONE MAN...
I don't believe I've ever, in like...15 years, written a RUclips comment. BUT, this is how you do an instructional video. Perfection.
Been subbed to you for a while now. Never stop saying "exactly" enough to guide people through 3D things with tact and poise. Good shit.
How to install stable difusion in 60 seconds.
Video Time : 43 seconds
well done
Can’t believe he cut down 43 seconds to 60
He built different. He mad the tutorial for the average joe
xDDD
actually with the 6 seconds intro and the fact that he already explained everything at second 30 makes the actual tutorial only 24 seconds long. 🤣
I search for this type video. Fast and straight forward. Sick n tired of people talking too much thinking they can make tons of money with long video. RUclips algorithm is changing. Thanks bro
This is a huge help, both the system and the excellent tutorial. All the details, none of the nonsense. You are a gift.
Thank God for you sir. You're the only one who's tutorials I can watch without getting distracted and forgetting what I'm supposed to be doing
Thank you for not dragging this out to 20 minutes like most people seem to do 😂
this is how i wish all tutorials were like, not too fast, not too slow and gets straight to the point, thanks!
Short, concise, everything we need and nothing we don't. Thanks!
Holy crap, way faster than the old way in your other video. I had to reinstall on a different drive and the last time I used that one. Thanks!
This is without a doubt the best tutorial for anything I've ever needed help with. No filler, just straight to the point. THANK YOU.
My freaking God. I literally spent 2 hours fighting around errors and dozens of long tutorials to get it work only to find out that this 40 sec video is the solution. THANK YOU
Thank you so much! i wanted to try installing SD for the longest time, but the long process of getting it looked really intimidating.
Not sure if you still read newer comments but thank you so much for this! I watched three other tutorials that failed but this worked flawlessly! Thank you!
Bro your tutorials are amazing. You deserve way more views!
Instruction unclear, ended up with a toaster in the bathtub. 💀
This has to be one of the most efficient tutorials on our planet.
Question, how do you update SD again? I completely forgot, lol.
Yeah, this could do with an installer - we are artists* at the end of the day!
*Well, its more like being soeone who commissions artwork and the tireless AI creates generally what you want
run the update.bat again
you are a legend brother, precise, direct, dont waste anyones time. Godspeed mate.
I am getting "ImportError: cannot import name 'packaging' from 'pkg_resources" Anyone know the way to fix this?
same problem here
Did you fix that?
@@alihp7844 go to the system file in your installed folder , then enter the python file(D:\AI\system\python) ,write cmd in the location bar (not the search bar but the one where it says the location youre in),press enter and write this command python -m pip install setuptools==69.5.1
Same problem
omg THANK YOU so much! I am so bummed with Playground AI eliminating their image generator, so needed an alternative. This helped me get up and running with stable diffusion quickly and easily!
The best tutorial I've ever seen. Thank you. Last time it took me 2 hour to understand how to install it.
This is a rare case where it's actually ok there's no time stamps. Thanks for a great tutorial.
ok thank you !
After install, how to run 'stable diffusion'???
0:27
@@musicenjoyer6026 It doesn't work
My command prompt seems stuck after running run.bat.. Surely that part shouldn't take more than a few minutes?
seconds for me.
@@dudleydooright make sure to press enter in the CMD if it is stuck
@@lefourbe5596 I meant it only took seconds for me. Worked fine.
Thank you for making these short videos! It's life-saving tutorials.
i have checkpoint error......not available or sort......did i miss something here? stable diffusion checkpoint dropdown is empty...
I keep clicking "run" it opens the window with the black background, then it does it's thing, finally it says "press any key to continue" when i press any key, it closes -- stable diffusion is nowhere to be found
I need help :(
I HAVE SAME CAN SOMEONE FIND A SOLUTION?
Make sure you enough space on your harddrive or ssd. It might not be downloading the files it needs to
@@SMSFRDROLE see my other reply
Thank you so much! My 2022 installation kept giving errors and I wasn't able to fix it, I just installed it fresh and it's working again =)
I got this error can you please help?
Python 3.10.6 (tags/v3.10.6:9c7b4bd, Aug 1 2022, 21:53:49) [MSC v.1932 64 bit (AMD64)]
Version: v1.8.0
Commit hash: bef51aed032c0aaa5cfd80445bc4cf0d85b408b5
Traceback (most recent call last):
File "C:\AI\sd.webui\webui\launch.py", line 48, in
main()
File "C:\AI\sd.webui\webui\launch.py", line 39, in main
prepare_environment()
File "C:\AI\sd.webui\webui\modules\launch_utils.py", line 386, in prepare_environment
raise RuntimeError(
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check
Press any key to continue . . .
This was though to my graphics card drivers not up to date. Thanks anyway
I wish all tutorials were like this, thank you!
I'm not incompetent to learn doing stuff the hard way when it comes to things that involve command prompt... but... boy was I really procrastinating on SD. Thanks for helping show that I didn't have to hack into the mainframe to get this onto my PC. xD
Subscribed. Mans went straight to the point and didn't try bullshitting his audience or spewing out the word "Free" when there is token systems and paywalls behind certain services.
I got this error
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check
Any way to fix this?
@@aberryth_dev I don't know you and I have no idea what's going in your life, but you matter bro please never forget this.I really mean it.Thank you so much for this detailed and easy to understand explanation, it helped me and now it works.I wish you all the best in life.
@@aberryth_dev Any other ideas? GPU driver up to date, I also tried to no avail with Python 3.9. The earlier more longer process gave me the same problem, I was able to get as far as this error when I tried to import.torch:
A module that was compiled using NumPy 1.x cannot be run in
NumPy 2.1.0 as it may crash. To support both 1.x and 2.x
versions of NumPy, modules must be compiled with NumPy 2.0.
Some module may need to rebuild instead e.g. with 'pybind11>=2.12'
I though Python 3.10.6 used the Numpy module version 1.x. ? I have no other versions of Python on this machine. Is the Geforce GT 430 just cursed?
@@rainbowrae267 it looks like the current version of numpy is too recent for the other dependencies.
First, try running update.bat (obvious, I know, but I’ve got to check)
Using the same method as above, try
-reinstall-numpy==1.26.4
Or if that doesn’t work:
-reinstall-numpy-upgrade
Or just -reinstall-numpy
I’ve never seen that before though, so honestly I don’t know
@@rainbowrae267 I think RUclips is deleting my replies for some reason
-reinstall-numpy==1.26.4
@@rainbowrae267 When I had that problem i had moved everything into the "Videos" folder on my computer. It was suggested in a reddit thread I found and it worked for whatever reason.
So try moving your "stable-diffusion-webui" folder, Python app and "Git-2.43.0-64-bit" app into your "Videos" folder. When I did that it worked
I install it and close it and then can't run,,,it said "site connection cant be reach ",,can you help me how to resolve ?
you have to launch run.bat from the folder everytime you want to use it
I did everything, once I clicked on "run" the black screen opened and it did it's stuff, then it said "press any key to continue" I pressed and the window closes.
How do i run stable diffusion?
ja frag ich mich auch
Make sure you enough space on your harddrive or ssd. It might not be downloading the files it needs to
@@thiloliedtke1971 see my other reply
do you have an nvidea graphics cards
i keep get that error :
ImportError: cannot import name 'packaging' from 'pkg_resources' (D:\sd.webui\system\python\lib\site-packages\pkg_resources\__init__.py)
any help please?
same issue
same as well ;_; but i still appreciate the uploaders brevity
edit: went into the webui folder of the extract and ran launch.py. this worked 4 me
I had the same problem, it occurs when setuptools version is 70.0.0. You must run the cmd on the system folder of webui and enter "python -m pip install setuptools==69.5.1".
@@0ee63 I tried it, but what do I run it on.
@@0ee63 Worked for me as well
Well done. That was quick and amazing :)
You are a God. Maybe a Semi-god...But a god..Thank you lord.
Didn't work for me. Webui.bat file says PYthon not found error. I don't know what to do
probably install python or just put the whole error into Chatgpt and give it some context. It will tell you what the error is and how to solve it
Hello, awesome video. Subscribed, because I like tech, cinematography, and vlogs. Keep it up! Peace man
i clicked on the Nvidia link sense i have an nvidia gpu like you said, but it keeps giving me the amd version
Can we add multiple models in Stable Diffusion local ?
Like example - an anime character's model with a cosplay clothing model and a background model.
the best you could do for that (i think) is make a background. first then img2img that use uh i forgot the name of the tool. its basically like a cutout the place you want the girl to be. then itll fill that
@@DeputyFish oh
@@SpideronOfficial its called inpaint.
it says if i run it i may be putting my pc at risk. I have winrar installed so should I just run it anyway or is it not a good idea?
always a risk but that just means that windows can't guarantee it's safety not that there is a particular vulnerability
best guide for anything ive seen in ages!
hey Royal Skies, thanks for the help. I tried both tutorials and none of them worked. The installion is sucessfull but the stable diffusion simply doesn't work. it gives all sort of errors when i try to generate anything, like this OutOfMemoryError: CUDA out of memory.
😔
CUDA is "memory" on graphic card for in a very very simple saying. Your GPU probably is not enough powerful to run it
How do we solve Cuda-out-of-memory? I don't know where to clear cache or reduce batch size.
This is gonna help me so much with drawing, I'll have some great reference images.
Instructions unclear, Now I have a velociraptor in my bathroom
as we say here in Brazil: "rápido e rasteiro". Thank You!!!!!!!!!!!!!!
Do I have to click run every time I use the program? Are there any ways to click open the program sir?
It says "Launching web with arguments" then press key to continue, its not giving me the link to start it why???
Could you do a tutorial on making clothing with physics on characters in unreal? Or link me to it if you already did..
did you ever find out the answer, please let me know
just found your channel, and this is the first video i saw, just what i needed, and right on time
question:when i try to download it, it gives me a error about torch, can you help?
I downloaded it but the system feels like its original horrible version where it couldnt make faces. Did I do something wrong?
same the faces are all blurry for me
same here... i used to have beautifulll stabe diffusions like a year ago, without any VAE etc i was just using anything v3 and it was beautifull but now the same checkpoint etc the same and everything is ugly... did you find solution?
@@ProfesjonalnyRUclipsr old comment so sorry for the necro but you'll need to get new models from civitai
@@jasonbernero9010 old comment so sorry for the necro but you'll need to get new models from civitai
keep getting the error "could not find a version that satisfies the requirement torch==2.1.2" how do i fix that?
Means that you might have to update python. or you need a lower python version. Just try to update python first if didn't work then try to install a lower version
Big fast Algo comment
btw, there's Stable Diffusion launcher called Stability Matrix. It acts like Steam, but for SD stuff. Pretty much 1-click solution.
It doesn't work. It says:
"ERROR: Could not install packages due to an OSError: [WinError 5] Access is denied
Consider using the ' -- user' option or check the permissions."
u made it look so easy but its taking me a lot longer because it took me 15minutes to figure out that i need to install a third party happ called git so it can use the git command, and then when u launch it it downloads dependancies like pytorch which in total is 2.5 gb and thats another 15minutes on my internet speed, also i had to run them as admin to get them to do anything so just over 30 minutes in total at this rate
Is it possible to create images in hexagonal shaped canvas in stable diffusion ? Please let me know how to do it.
yes and no. You sill will have a white (or w/e color border) around the hexagonal shape generation. The final image will have 4 sides, with a hexagonal shape interior generated image. There are different ways to do this. One way in inpainting.
@@highcommander2007 As you mentioned regarding inpainting. Can you give an example prompt to create image on a hexagon shaped canvas . Your reply will be greatly appreciated. Thanks in advance.
I installed this, couldn't get it to work at all. How do I remove it from my pc? I can't see trace of it in the "uninstall program" tab
just delete it. it's not installed it's literally just downloaded file
Best youtube tutorial ever
When I try to generate any image, even a simple one, the computer freezes. And only physically turning off the computer to unfreeze it.
tell me your specs
How do i know its safe? It needs to permit access to the system32 commands. Seems sketchy
40 sec speeeereeed 😂. Handsome.
Btw comfyUi is even faster to install coz it doesn't need the upgrade part.
The trainer tool would be harder
So to open it do we just use the run file?
your are a legend. Other tutorials make me very aggresive
I'm getting this error, when trying to do it. FileNotFoundError: No checkpoints found. When searching for checkpoints, looked at: - file C:\Ai Art\sd.webui\webui\model.ckpt - directory C:\Ai Art\sd.webui\webui\models\Stable-diffusionCan't run without a checkpoint. Find and place a .ckpt or .safetensors file into any of those locations.
Time taken: 0.1 sec.
Is there a way to add other models when it runs?
hello i keep getting this error. i followed your tutorial exactly. how do i fix this.
NansException: A tensor with all NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check.
Just ask AI my brutha.. :D
The error message you provided appears to be related to a software exception in a program or library called Unet. Unet is commonly used in the context of deep learning, particularly in image segmentation tasks.
The error message indicates that a tensor (a multi-dimensional array) containing NaNs (Not-a-Number) values was produced during the execution of the Unet model. NaNs typically occur as a result of undefined mathematical operations, such as division by zero or taking the square root of a negative number.
The message suggests a few potential solutions:
Set "Upcast cross attention layer to float32" option: If there is an option in the settings of Unet related to upcasting cross attention layers to float32 (single precision floating point numbers), enabling this option might resolve the issue.
Use the --no-half command-line argument: This command-line argument suggests disabling the use of half-precision floating-point numbers (float16), which may not be supported by your hardware.
Use the --disable-nan-check command-line argument: This command-line argument suggests disabling the check for NaN values altogether. However, this might not be a recommended solution as it could potentially mask underlying issues in the model or data.
Consider trying these suggested solutions based on your specific use case and requirements. If you're unsure about how to implement these solutions or if the issue persists, you might need to seek further assistance from the developers of Unet or from a community/forum where Unet users discuss and troubleshoot similar issues.
u deserves more subs
Subscribed. This video...fantastic. Thanks man.
Straight to the point! Love it
It installed 1.5, is there a way to add newer versions as well?
⭐⭐⭐⭐⭐
This guy needs to do a Ted Talk on RUclips video tutorial making. Didn't even waste a second to ask the viewer to subscribe or like. I subscribed to the channel and liked the video simply because of that fact🤪
Why i have 1.10.1.... its not 1.5. Or is it higher than 1.5?
A tensor with all NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check.
anyone know how to fix this
I try to run it but nothing happens.
holy shit, that was fast. thank you
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check
How can I fix the problem?
If you are having trouble installing Stable Diffusion, try stopping Windows updates in the background in the task manager. Once you have done that, try running the installation again.
@@CollaborativeMindAI It doesn't work.
Hey, someone can help me ? I got that issue :
AttributeError: 'NoneType' object has no attribute 'lowvram'
If someone can solve that please
Hello RUclips reader.
I following the video and got the following message "RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check
Press any key to continue . . ."
I have a Intel UHD Graphics 630
Please Help.
Did you download the NVidia version for intel?
@@samp6162 how to download intel version? can u inform me about that ? i can not see the "beta" text anywhere
@@samp6162 sorry for the late reply. I downloaded the Intel version.
Hey man, sorry to interrupt. did u manage to solve this? Im having the same situation@@shawnwykoff8744
dont work for me it says AttributeError: 'NoneType' object has no attribute 'lowvram' when i wanna make a pic
love you bro as always
Best tutorial ever seen
--cuda-malloc bunu eklememi öneriyor nereye yazacağız. webui bat dosyasına olmuyor denilmiş.
hmm mie doesn't have a cpkt. what should i do?
God-tier video, thanks!
Yeah, but what about us Mac people?
Getting ModuleNotFoundError: No module named '_socket'. Please help!!
doesnt work do i need python too?
how do I improve the image quality?
I'm having trouble between "run" & "ya done."
I keep clicking "run" it opens the window with the black background, then it does it's thing, finally it says "press any key to continue" when i press any key, it closes -- stable diffusion is nowhere to be found
I need help :(
Did you ever fix it???
how do i know if the "automatic1111"is trusted ?
apparently its not from official
Easy. Run it. If your computer gets taken over by viruses, it's not safe.
the astronomical number of star on automatic1111 that you can look up on google is a nice hint.
128k star as of today.
if that can't be trusted on the official github website i don't know what is.
is there a way to install 1.5 like this?
been waiting for 30 minutes for it to ''install requirements''. Is that normal?
I had to re-run it and it worked, tho, it says now: ''press any key to continue''. I have pressed a key couple minutes ago and nothing is happening.