I have success with Easy Diffusion. Stable Diffusion? Tried over 20 times with other vids, so I'm dubious. If this tut actually tames the beast, I'll be sure to let you know...
bruh this what mine says C:\stable-diffusion\stable-diffusion-webui>git pull Already up to date. venv "C:\stable-diffusion\stable-diffusion-webui\venv\Scripts\Python.exe" Python 3.10.6 (tags/v3.10.6:9c7b4bd, Aug 1 2022, 21:53:49) [MSC v.1932 64 bit (AMD64)] Version: v1.4.0 Commit hash: 394ffa7b0a7fff3ec484bcd084e673a8b301ccc8 Installing torch and torchvision C:\stable-diffusion\stable-diffusion-webui\venv\Scripts\python.exe: No module named pip Traceback (most recent call last): File "C:\stable-diffusion\stable-diffusion-webui\launch.py", line 38, in main() File "C:\stable-diffusion\stable-diffusion-webui\launch.py", line 29, in main prepare_environment() File "C:\stable-diffusion\stable-diffusion-webui\modules\launch_utils.py", line 265, in prepare_environment run(f'"{python}" -m {torch_command}', "Installing torch and torchvision", "Couldn't install torch", live=True) File "C:\stable-diffusion\stable-diffusion-webui\modules\launch_utils.py", line 107, in run raise RuntimeError(" ".join(error_bits)) RuntimeError: Couldn't install torch. Command: "C:\stable-diffusion\stable-diffusion-webui\venv\Scripts\python.exe" -m pip install torch==2.0.1 torchvision==0.15.2 --extra-index-url download.pytorch.org/whl/cu118 Error code: 1
Excellent presentation. No stupid music, no idiotic graphics or cutaways to film clips, just calm instructions from someone who knows what they're doing. Well done.
Gosh, I'm a 60 year old grandmom who is a techie that can run circles around most young people I know - at least in my businesses. I installed and manage all the smart home devices in my 9500 sq ft home. I installed and manage all the technology in my businesses and I like to use the latest devices so I'm constantly upgrading. Not all old people are technically challenged.
Thanks you Seb 🙏🏽 I am also named Sebastian, and over 60 years old, so this was rather intimidating for me until I came across your wonderful tutorial where you so calmly and systematically teach us the minimal items we need to get started with Stable Diffusion.
You are the only person I have seen explain calmly so that others understand, you are a wonderful person because you teach us beginners calmly and step by step
the problem is that the extensions and even the programs are being updated too fast ... example Comfyui i find this: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-CWKVhWHVecg.html
Just a tip for anyone finding trouble saving the batch file, just reinstall the a1111 folder to desktop, it's usually because you need administrator permission to download and save things in program files 👍👍👍
I want to compliment you on making the best tutorial regarding SD/Auto1111 installation that I've seen. I don't know why it's so hard for people to be concise, clear and not rushed when making instructional videos, but for some reason it's a common issue. I wish I had watched this when I was originally installing, the amount of perplexing explanations I have sat through is laughable. So again, well done, and thankyou ever so much for your work.
I was going to make a video like this simply because of how many people have asked me how to install and get Stable Diffusion running, and I suck at explaining things solely through text. Thank you for saving me the time. Very good, comprehensive explanation. Just as much as you need and easy to understand.
Great vid, helped a lot. Small note, youtube's target loudness is -14 LUFS. I havent seen your other videos so it may have been fixed since then, but if not, thats the reason your video is so quiet compared to most others on the platform. Thanks for the tutorial!
he is trying too hard to be "ASMR" and it completely ruins the video. All his voice is doing, talking so unnaturally, is creating constant a low-tone reverb which is migraine inducing. FFS
Couldn't be easier to follow. No "wait - what? that's not what I'm seeing on my screen" moments. Custom installing anything for non-coders or the techincally proficient is always intimidating, but this tutorial makes you feel like you're in safe hands. Thank you!
I'm an AI Dunce. I just started using it recently and this is BY FAR the most useful (new user friendly) video I've found. Thank you SO MUCH for helping students learn the material, rather than trying to impress viewers with your existing knowledge. You're a gentleman and a scholar. Thank you.
Wonderful video. Very clear and concise. The way you speak reminds me of when I'm working from home on a call and the baby's asleep in the next room. Cheers!
I didn't know about the styles.csv! What a timesaver, I'll probably add some of my own styles to it later on. Thanks for this! I'll also share this video with anyone curious about creating SD images for themselves as your calm and informative explanation was excellent.
I want to thank you for this amazing tutorial. However I could not run it and in CMD there was some error related to CUDA. Do you have any suggestions for that? @@sebastiankamph
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.
Hii bro add this argument in Commanline_ARGS- --medvram --xformers --autolaunch --no-half and save it and then run the batch file again and it will start (note after you generate one image exit the page and reload it or click on batch file again after every one generation)
Thanks Sebastian, I finally managed to get Stable Diffusion working because of your great video. Now I just need to master it... So I'll watch more of your content for sure. Subscribed.
PSA: I got a permission error while trying to clone, I had my SD folder inside of my program files folder, which didn't have permission. I made a new folder in C and it worked perfectly. Thanks for the video, it was a great help!
my installation of A1111 crapped the bed so I decided I should just scrap everything and start from scratch, luckily still have python and git sitting on my desktop to skip some steps. seems a lot more straightforward than the first time I installed A1111 a while ago.
Great tuto, man. I had a error with torch download at the end but I just install a11111 folder on C: and everything is working fine. I think the problem was because I installed first a11111 on my SSD instead my main drive. Thanks again!
@@Cookieheart308 You need to do everything what he said on the same drive. My mistake was to install the a11111 on my SSD instead of my principal drive.
that and the python version matters, make 100% sure it's 3.10.6 if not you'll also need to delete a folder called "venv" if it appears in your SD root folder before reinstalling python
Thank you so much! I came for the x-formers installation part, but I'm very glad to have found (and that you made it available) your style.csv file. I went from an 18 minutes image generation prior to x-formers to... 33 seconds after installation (a 640x768 with high res x2, using a GTX 3060 12GB).
I'm getting an error code 1 after launching webui.user. ERROR: Could not find a version that satisfies the requirement torch==2.0.1 (from versions: none) ERROR: No matching distribution found for torch==2.0.1
Very nice. Just wanna add that, @5:41 where we're adding COMMANLINE_ARGS, we'll also want to add --theme dark for anyone who's like me and uses a browser that defaults to that white-screen nonsense.
i have 4GB VRAM, then I was tryed improve to --lowvram, and the result is super ugly, but from this video it's ok to be --medvram, now the result is fine. thanks man, I subscribing ur YT now sorry for my bad english
Thanks for the tutorial, Sebastian, it worked like a charm. I tried to install Disco Diffusion into my WSL2 a few months back and failed miserably, so I was massively surprised when this worked literally without any hiccups.
While the explanaion seems good, unfortunatelly i'm one of many people for whom the installation didn't go as smoothly I had for example to add --skip-torch-cuda-test to the cmd and i still can't generate a thing (nothing happens). I'm not sure if you mentioned the requierments . I think it would be good to know what is expected (for example I have Intel not NVIDIA and it makes a difference. The choice of browser seems to make a difference too. And so on. I'm sorry if you actually mentioned any of those or told in the video where to look for more information. But I had to add this comment to convey that one might run into some problems while installing as suggested
Best video out there for a beginner like me to understand stable diffusion and automatic 1111. Thank you very much. Now I just need to see how to activate the Open VINO for intel integrated graphics.
So, I just got the "can't install torch,.." and "Couldn't find,..." Errors. Which is strange, since I have the correct Python installed. I dunno what I am doing wrong.😅
Just let the command line continue. When the Web ui doesn't open up run the batch file one folder deeper webui.bat. If error persist Google "stable Diffusion can't install torch" the git issues have some comments that detail how to fix it
So you have to hit Start, > "Manage App Execution Aliases" > turn off both of the ones that say Python.. but now my problem is I says I failed to install torch.. sigh..
Hi Sebastian just to let you know... I had to add in note pad for the batch file set PYTHON=py otherwise it wasn't working. keep telling me pyton wasn't found.
HI, Sebastian. I am excited and curious about generative AI and found your videos to be greatly valuable. Content is exactly what I am looking for and the quality is just perfect. Thank you so much!
I'm on AMD 6800XT and Windows 10. I did some tests with a fresh install (a week ago) and found the optimal launch settings: . --opt-sub-quad-attention --medvram --no-half --no-half-vae --precision full . they give a 20-30% speed boost compared to the following: . --opt-sub-quad-attention --medvram --no-half-vae . --opt-sub-quad-attention - increases the maximum possible resolution to 1392x1392, compared to the deprecated --opt-split-attention - 1024x1024. . After several months of working with A1111 on Linux, I installed it on Windows a week ago, which I am very happy about! During these couple of months, I managed to hate Linux with all my heart.
You're very welcome, and thank you kindly! Your support means the world to me 😘 Feel free to join our Discord and message me there if you want vip channel chat access.
You can use sdp now instead, correct. It's almost the same speed and deterministic. Some tests have shown xformers to be marginally faster, however not deterministic. I'm sure there are settings when t2 and sdp come out on top though.
when i try to launch webui-user it says Couldnt launch python, exit code: 9009 stderr: Launch unsuccessful. Exiting. Press any key to continue... I did everything as u did step by step... any advice ?
I’ve never touched python in my life and I was able to successfully follow this and install it. I even have it running on an older laptop using my CPU. (Since I don’t have a gaming pc) First came across an error of 128 and I figured out from searching that where you edit in the web ui in notepad to add: -skip-torch-cuda-test Then I opened it again, got a new error, but I ignored it, opened it again, and it worked!!! 🎉
I need to generate pictures for academic purposes using Stable Diffusion, Automatic 1111, and MidJourney. My laptop specs are: Lenovo ThinkPad X1 Yoga Windows 10 Pro Intel Core i7-7600U @ 2.80GHz 16GB RAM Intel HD Graphics 620 341GB free on 476GB storage Can my laptop handle these platforms effectively? Any advice?
I just followed this tutorial. You don't need to add --autolaunch anymore to open the web interface automatically after you launch automatic1111. That just happens by default. If you want to turn that off, you do that under settings -> system in the web interface.
I like you're calm way of using your voice, it's a welcome relief in comparison to the hysterical insanity I hear from my kids youtube channels in a cacophony of different idiots shouting their guts out spouting nonsense in my room, someone please tell me how to stop this without having to be a clown all day to entertain my kids who forgot to be kids. I know I am. Complicit, it sneaked up on me during the lockdowns, letting go of screentime seemed like the only way to remain sane. Excuse me for going very offtopic here..but hey you got a new subscriber
Thank you kindly! I understand the sentiment and try to provide something else in this crazy world of ours. It appeals to a smaller audience, but I know it's valuable to those people. Thanks for reminding me of that 🌟
ME TOO WHAT A FANTASTIC VIDEO UPLOAD, Thank you very much, Sebastian. I truly appreciate you and your thorough instructions….and your serene manner. 😊❤👍👍👍
You are amazing. Thanks for the clear & concise jumpstart. I'm going to watch all your related stable diffusion videos. I have butterflies just thinking about the possibilities ahead. Yeeeeeee Ha! :)
If you're working on a subpar system that can still somehow run SD (if you're running in s/it and not it/s), I'd recommend disabling the live preview. As much as it's cool to watch, disabling it improved performance on my system by almost twofold and it also forces me to go do something else while I run my prompts because I was kind of addicted to seeing the image evolve on my screen which quickly becomes a huge waste of time when it takes several minutes per image. Thanks for the tutorial! I have to re-install A1111 every few months for reasons unknown and it seems to have gotten simpler than it was last winter!
My preference for install is using WSL. I use Ubuntu 22.04 currently and I have Automatic1111 installed. I have found almost no channels that cover Linux but following instructions on the Git page works well. Once installed, everything else should be universal across any OS.
If you're running under Windows 11, after installing Python and confirming it's in your PATH (add if it's not), remember to disable the "App execution aliases" in Windows or it won't find Python 3.10.6 where it was installed.