Easiest Way to Install & Run Stable Diffusion Web UI on PC by Using Open Source Automatic Installer

[FurkanGozukara]

Easiest Way to Install & Run Stable Diffusion Web UI on PC by Using Open Source Automatic Installer

Full tutorial: https://www.youtube.com/watch?v=AZg6vzWHOTA

Our Discord : https://discord.gg/HbqgGaZVmr. How to download and install Stable Diffusion Automatic1111 Web UI interface. If I have been of assistance to you and you would like to show your support for my work, please consider becoming a patron on 🥰 https://www.patreon.com/SECourses

Playlist of Stable Diffusion Tutorials, Automatic1111 and Google Colab Guides, DreamBooth, Textual Inversion / Embedding, LoRA, AI Upscaling, Pix2Pix, Img2Img:

https://www.youtube.com/playlist?list=PL_pbwdIyffsmclLl0O144nQRnezKlNdx3

In this video I am explaining how to install Stable Diffusion famous #Automatic1111 Web UI on a PC with the easiest possible way by using an open source installer. This installer is made by a Leo Mozoloa (https://twitter.com/leomozoloa) who is an active community member.

The Github repo of the automatic installer we used in the video : https://github.com/EmpireMediaScience/A1111-Web-UI-Installer

The exe release version 1.3.2 I used in the video: https://github.com/EmpireMediaScience/A1111-Web-UI-Installer/releases/download/V1.3.2/A1111.Web.UI.Autoinstaller.v1.3.2.exe

The Github repo of the official Automatic1111 #StableDiffusion #WebUI : https://github.com/AUTOMATIC1111/stable-diffusion-webui

Python 3.10.8 version direct link x64 : https://www.python.org/ftp/python/3.10.8/python-3.10.8-amd64.exe

Stable Diffusion official 1.5 model file direct link : https://huggingface.co/runwayml/stable-diffusion-v1-5/blob/main/v1-5-pruned.ckpt

00:00:00 Intro - we are going to show 2 ways to install web UI on the PC

00:00:28 Developer of the Automatic Installer of the Stable Diffusion Web UI

00:00:31 Explanation of why we can trust this installer

00:00:58 From where how to download this automatic installer

00:01:26 How to start installation of the automatic installer. Run as administrator is the key

00:02:08 Installer exe beginning screen and we start to install

00:02:28 How you can find our Discord link to join and ask any questions

00:03:25 Where you can find more information related to different Stable Diffusion models and vae files

00:03:50 Pick the Web UI installation folder and start installation

00:05:29 How to send shortcut of installed launcher to the desktop and start the Stable Diffusion Web UI application

00:06:06 The launcher interface and settings

00:07:23 Which another video you must watch to understand how to use Stable Diffusion Web UI better

00:09:16 Web interface is started and we are doing first image generation

00:10:46 Entering prompt to start image generation

00:11:10 How to upscale an image by using AI - awesome quality

00:12:04 Starting manual installation of the Stable Diffusion Web UI from its GitHub folder

00:14:48 If you get error during Python installation, another version of this product already installed

00:15:55 How to verify installed and actively selected Python version

00:16:19 How to change default Python from environment variables path variable

00:17:58 How to run Web UI from manually installed folder after installing Python

00:18:44 Fixing the encountered error during running the Web UI

00:20:15 How to install downloaded stable diffusion models

00:21:17 How to open started Web UI application

Of course using Stable Diffusion on your computer is the best and the most convenient way. However remember that, this model requires a decent graphic card. Therefore, if you don't have such graphic card, check out our other videos that we demonstrate on Google Colab notebook which runs on the cloud and requires no graphic card.

Stable Diffusion: The Best AI Image Generation Model Yet?

In the world of artificial intelligence and machine learning, image generation models have become increasingly sophisticated and powerful. These models are used for a variety of purposes, including creating realistic images for use in film, video games, and other media, or for generating novel or creative images for use in art, design, and other creative fields.

Among the many AI image generation models that have been developed, one stands out as being particularly impressive: Stable Diffusion. According to its developers, Stable Diffusion is the best ever generated AI image generation model.

If you want to use the power of Stable Diffusion for yourself, you'll be glad to know that there is an easy and convenient way to do so: the web UI for the Stable Diffusion model. This user-friendly application makes it easy to access and use the Stable Diffusion model from any computer with an internet connection, without the need for any special software or technical expertise.

The web UI for Stable Diffusion is designed to be as intuitive and user-friendly as possible, with a clear and straightforward interface that makes it easy to upload images, adjust settings, and generate new images. Whether you are a professional artist or designer looking to create high-quality images, or a researcher or developer interested in exploring the latest advances in machine learning and AI, the web UI for Stable Diffusion is an essential tool that you won't want to be without.

Video Transcription

• 00:00:01 Greetings everyone.

• 00:00:02 In this video, I am going to show you how you can install, run and use famous Stable

• 00:00:07 Diffusion web UI developed by Automatic1111 with two different ways.

• 00:00:12 The first way I am going to show is using an open source installer exe just like in

• 00:00:16 traditional applications we install.

• 00:00:19 The second way will be manual installation of the web UI.

• 00:00:22 This exe is developed by one of the Stable Diffusion community members Leo Mozoloa as

• 00:00:28 can be seen here, this is his profile.

• 00:00:31 Usually I do not trust any exes, however, this one is an open source application and

• 00:00:36 you can compile it from the source code.

• 00:00:38 You can also check all of the source code on the GitHub repository.

• 00:00:41 This link I will put it in the description of the video and we can go inside the folders

• 00:00:47 and see each one of the code it contains.

• 00:00:53 Therefore in this case, we can trust it.

• 00:00:55 So let's begin with downloading the installer exe.

• 00:00:58 So where do we download it?

• 00:00:59 We download it from the releases.

• 00:01:02 The releases are displayed at the right as you can see here with my mouse.

• 00:01:05 When I click it, I can see all of the releases.

• 00:01:09 Just click the exe and it will start downloading.

• 00:01:14 By the way, you may get a message like this.

• 00:01:16 This exe is not commonly downloaded and may be dangerous.

• 00:01:20 If you see that just right click here and click.

• 00:01:24 Keep OK.

• 00:01:25 After the download has been completed, just go to your downloads folder.

• 00:01:29 Right click the exe and run as administrator.

• 00:01:32 It is important to run as administrator because this exe will download and install files.

• 00:01:40 Don't worry about that because it is hosted on the GitHub repository.

• 00:01:46 Also this application is now pinned by moderators of the stable diffusion community and also

• 00:01:51 I have compiled myself this exe by using advanced installer application and I have verified

• 00:01:59 it.

• 00:02:00 This application you can also download trial and try, but it is not necessary for you.

• 00:02:05 OK, now this is the installer exe beginning screen.

• 00:02:08 We need to install Python and this exe will do that as well.

• 00:02:13 So you see there is GIT and Python.

• 00:02:15 I click next.

• 00:02:18 It will do everything automatically for us.

• 00:02:22 If you encounter any problem during following this tutorial, please go to our page YouTube

• 00:02:29 page.

• 00:02:30 Click about and here you will see our Discord channel.

• 00:02:35 Just click it and join our Discord channel and ask me any questions, any problems that

• 00:02:39 you encounter.

• 00:02:41 The Python installation may take some time, so be patient with that.

• 00:02:47 OK, so the Python installation and the git installation has been completed, so there

• 00:02:53 are several options.

• 00:02:55 You should pick clean install if you are going to override your existing or previous install.

• 00:03:01 You should also click this box.

• 00:03:04 Install SD checkpoint 1.5, version 1.5 and the newest vae.

• 00:03:12 These are, I am explaining what are these in our another video.

• 00:03:18 So if you are wondering what are these version 1.5 and new vae you can go to our channel.

• 00:03:25 Click playlist and in here you will see stable diffusion dreambooth click that playlist and

• 00:03:31 in here you see we have how to use SD 2.1 custom models on Google Colab or we have how

• 00:03:38 to use stable diffusion 2.1 and different models in Web UI.

• 00:03:42 This video actually will explain to you a lot regarding what are these.

• 00:03:48 OK, so now I am going to pick the folder where I want to install.

• 00:03:52 I want to install it inside my C drive.

• 00:03:56 So I am clicking the three dots here to pick the folder.

• 00:04:00 I am clicking C and in here I am clicking select.

• 00:04:04 You see now it is selected in the root folder, but I need to have another folder.

• 00:04:10 So I am going to type like a1111WebUI.

• 00:04:15 OK and then I am clicking I understand and click install.

• 00:04:22 Now it will also download the necessary files for me and we don't need to do anything manually.

• 00:04:30 During the download, it doesn't show you any messages.

• 00:04:34 However, if you open your task manager with right click and click task manager like here,

• 00:04:40 you can go to ethernet and you can see how much download your computer is doing.

• 00:04:46 As you see, currently I am downloading like 100 megabits, which is my maximum bandwidth.

• 00:04:53 OK, after the file files has been downloaded, it will quickly install them and it will open

• 00:05:00 the installed folder.

• 00:05:03 As you can see, the folder is like this.

• 00:05:05 Now we have launcher generated by our installer and we have the GitHub folder of the official

• 00:05:14 Web UI.

• 00:05:15 It is this one and it has automatically downloaded it for us and set it up.

• 00:05:21 And we have the launcher shortcut icon.

• 00:05:25 You can also pin this to taskbar, send it to your desktop, whatever you do.

• 00:05:30 For example, I will send it to my desktop, send to desktop.

• 00:05:34 And now in my desktop I have it and it has been installed.

• 00:05:42 As you can see now, I am clicking finish and now we can run it from directly from the launcher.

• 00:05:49 It is asking me asking me to do I want to launch the from its shell command?

• 00:05:57 And I said yes.

• 00:05:59 And now the shell command has been run and the automatic Web UI launcher has been started

• 00:06:06 so we can do automatic update.

• 00:06:09 As you can see, the arguments are getting setted it in the left.

• 00:06:13 Automatic update extensions, clear generated images.

• 00:06:15 You shouldn't pick this if you if you know what you if you're not sure what is this doing.

• 00:06:21 Low VRAM.

• 00:06:22 If your graphic card is lower than 8GB, then you should check this.

• 00:06:27 I have RTX 3060.

• 00:06:32 The RTX 3060 is a super card, super card for machine learning tasks.

• 00:06:39 And I already have a video regarding that on my channel.

• 00:06:44 If you wonder how powerful it is, you can check that video, this one and read the comments

• 00:06:53 that people made.

• 00:06:56 You should also click enable xformers.

• 00:06:59 This enable xformers will improve your speed if your graphic card is supporting that.

• 00:07:08 There are also optional checkpoint folders and default vae files.

• 00:07:12 If you don't know this yet, you shouldn't change them.

• 00:07:17 But if you want to learn more about them, just go to our page, click videos.

• 00:07:23 And from here, find how to use stable diffusion version 2.1 and different models in the web

• 00:07:29 UI video.

• 00:07:30 This video will help you tremendously to understand them.

• 00:07:33 I will also put a link of this into the description of the video.

• 00:07:37 There are also additional launch options, and you can also type them.

• 00:07:43 Currently only this one is typed.

• 00:07:47 You should also type here.

• 00:07:49 --no-half.

• 00:07:50 This will make it work with stable diffusion version 2.1.

• 00:07:57 And I am also explaining that in this video as well, how to use stable diffusion 2.1.

• 00:08:03 And that's it.

• 00:08:04 That's all we need to start.

• 00:08:07 And then we click launch web UI.

• 00:08:10 It is generating a virtual environment for Python.

• 00:08:17 This is only first time initialization and you won't need it.

• 00:08:21 So you see it has installed Python 3.10 version.

• 00:08:28 This is one of the supported versions by the web UI 3.10.8.

• 00:08:34 If you have different Python installations in your computer, it is fine.

• 00:08:39 It will use the latest installed one because I am assuming that the web this installer,

• 00:08:47 installer will set the default Python path in the environment variables.

• 00:08:53 Now we are waiting it to install necessary other files, for example, torch and torch

• 00:08:59 vision.

• 00:09:00 These are necessary to run the stable diffusion model and generate images.

• 00:09:07 OK, so the installation has been completed.

• 00:09:11 You see, I don't have any error messages.

• 00:09:13 This is really important.

• 00:09:16 And then it will automatically start the web interface that we can do image generation

• 00:09:24 and image manipulation with stable diffusion.

• 00:09:27 OK, by the way, you need a decent graphic card to make it work with many of the features

• 00:09:34 of this AI model.

• 00:09:38 If you don't have such graphic card, then on our channel, you can go to playlists.

• 00:09:45 And in here, click the stable diffusion and watch these transform your selfie and how

• 00:09:53 to use SD 2.1 custom models on Google Colab.

• 00:09:57 Both of the videos explaining how to use stable diffusion on a Google Colab notebook, which

• 00:10:06 uses the GPU provided by the Google.

• 00:10:10 So you don't need a strong GPU.

• 00:10:12 However, in this tutorial, we are going to use our own computer's GPU.

• 00:10:19 As you can see, my GPU is already being used a lot, but it is fine.

• 00:10:24 Now everything is ready.

• 00:10:26 First, we need to pick the model that we want to use, which is this one automatically downloaded

• 00:10:33 and got ready for us.

• 00:10:34 Now we can enter prompts and generate images and use all of all other features.

• 00:10:43 So in this video, I will just generate one image.

• 00:10:46 Okay, I have entered my prompt as like this and my negative prompt.

• 00:10:51 I have selected the sampling sampler as DPM++ SDE Karras and other options are like default.

• 00:11:00 If you want to learn more about this, please watch the other videos in the channel because

• 00:11:06 these are all different things.

• 00:11:10 I will also show one another thing.

• 00:11:12 Just click this image and send to extras and in here you can upscale it.

• 00:11:17 The one of the best upscaling algorithm I have found is R-ESRGAN 4x+ and let's upscale

• 00:11:26 it to three times and just click generate.

• 00:11:29 This will also download another necessary model file for R-ESRGAN 4x+.

• 00:11:38 It is not very big as you can see.

• 00:11:40 Then you will see it is generating the upscaled image and it is ready.

• 00:11:45 When we click it now you see it is upscaled.

• 00:11:49 Okay, so this is all for automatic installation with using the installer.

• 00:11:55 Now I will show how you can install and run everything manually.

• 00:12:03 Okay.

• 00:12:04 The manual installation is tedious as you can imagine.

• 00:12:07 First we are opening the GitHub repository of the stable diffusion web UI.

• 00:12:12 I will put the link in the description.

• 00:12:15 Then click code and from here click local and from here download as zip.

• 00:12:20 This will download the entire folder and it is it has been downloaded pretty fast.

• 00:12:26 Cut it and then go to the folder that you want to.

• 00:12:30 You want to install it.

• 00:12:31 So I will generate a new folder in C drive and let's say webUI2 okay and paste it in

• 00:12:42 there.

• 00:12:44 And in here right click and extract here.

• 00:12:47 This is a zip file.

• 00:12:49 Okay I will cut and everything inside inside here and I will paste them inside here.

• 00:12:56 Okay by the way, whenever the source code gets updated in this repository, you have

• 00:13:04 to manually update it.

• 00:13:06 It won't be updated automatically.

• 00:13:10 Okay so to make this work, we need to have installed python.

• 00:13:15 So how we are going to install python.

• 00:13:18 The suggested python version is python 3.10.6 and when you click it it will open all of

• 00:13:26 the python versions.

• 00:13:27 Do not download the latest python because it won't work.

• 00:13:32 Some of the things are not compatible with the latest version.

• 00:13:37 So from here we go to we are going to search control-F 3.10.6 here and from here we go

• 00:13:54 to files and from here click windows installer 64 bit it will get installed.

• 00:14:00 Currently I have this installed but we can reinstall it.

• 00:14:04 I am going to show in folder right.

• 00:14:06 Click and run as administrator.

• 00:14:10 Okay and in here you see there are install now and customize installation and add python

• 00:14:17 to add python to path.

• 00:14:19 This is important.

• 00:14:20 If you don't make this, it won't work.

• 00:14:23 Okay I will click customization.

• 00:14:25 Everything is selected next.

• 00:14:27 Then you can pick the folder where you want to install.

• 00:14:30 I want to install it in C drive with version 2 like this and you can also click precompile.

• 00:14:40 Okay this should be enough sufficient.

• 00:14:42 Click install.

• 00:14:43 Okay you see I have an error because another version of this product is already installed.

• 00:14:49 Okay I am uninstalling the pythons version 3.11.0 and 3.10.8.

• 00:14:58 This is the one that has been installed by the automatic installer and then we will manually

• 00:15:04 install this one to demonstrate.

• 00:15:09 Okay both python versions has been uninstalled, so let's install 3.10.8 version.

• 00:15:17 Okay which is here I am clicking it I am clicking the download.

• 00:15:21 I will also put a link of this into the description show in folder right click run as administrator.

• 00:15:29 Okay you add python exe to path customize installation.

• 00:15:34 Next, let's pick the folder to install.

• 00:15:37 Okay let's name it like this: precompile.

• 00:15:41 Okay install.

• 00:15:42 Okay now getting installed.

• 00:15:45 Hopefully it won't cause any errors.

• 00:15:53 Okay the python has been installed successfully.

• 00:15:55 Let's first verify it.

• 00:15:58 Open a command line window with typing cmd.

• 00:16:02 Like this command, prompt, enter and in here just type python.

• 00:16:08 Okay you see my default python is 3.9.9 therefore, it may not work and it will not work actually.

• 00:16:19 So I have to fix it from environment variables.

• 00:16:22 If you if you didn't have any installed python, then it should automatically work for you.

• 00:16:29 So I am going to add I'm going to add it to my edit path and in here I am going to move

• 00:16:38 I am going to add my new installation.

• 00:16:41 Actually it is on the top so probably I need to change the path here.

• 00:16:47 Okay let's see the newest installation is not here, so click new and from here.

• 00:16:53 Click browse in here.

• 00:16:55 Go to C drive where we installed it.

• 00:16:58 We installed it in let's see python.

• 00:17:02 Yeah, in here first selected.

• 00:17:06 Move it to the top.

• 00:17:09 Okay then I will just edit this actually the bottom one.

• 00:17:20 Okay for editing, click edit and click browse and go.

• 00:17:27 Okay in here python and in here I'm selecting scripts okay, click okay, click okay and you

• 00:17:37 need to open another cmd command prompt window type python.

• 00:17:43 Okay now you see my default python is 3.10.8.

• 00:17:51 Now we are ready to move to the next step.

• 00:17:55 So I am entering the folder where I have extracted the GitHub files and in here first I need

• 00:18:02 to edit web user.bat file.

• 00:18:05 Right click it and click edit I need to add 2 command line arguments.

• 00:18:11 The arguments we need to add are --xformers space character and --no-half and then click

• 00:18:21 save.

• 00:18:22 Okay now we are ready to launch the interface.

• 00:18:26 We are going to webui-user.bat and it will generate first a virtual environment as same

• 00:18:34 as in the previous methodology.

• 00:18:38 Okay the installation has been completed, but we have an error.

• 00:18:44 The error is that it says that we don't have any model to run the stable diffusion.

• 00:18:50 So first we need to download and install our model.

• 00:18:56 As I have already said multiple times, I am fully explaining different models and how

• 00:19:03 to use them with the web ui in this tutorial video.

• 00:19:09 So from here I am going to click official repository actually not official repository.

• 00:19:18 So let's download not the official model, but let's download anything v3 model: this

• 00:19:26 is a very famous model and from here I am going to click anything v3.

• 00:19:34 I will use the pruned file.

• 00:19:36 Okay, just download it alternatively, you can also download the official version just

• 00:19:44 click stable diffusion.

• 00:19:46 Actually, this could be hard.

• 00:19:48 So in google, just type stable diffusion version 1.5 if you type it to the google, it will

• 00:19:57 bring you the hugging face page.

• 00:19:59 From here and from here click files and versions and from here click v1-5-pruned.ckpt and from

• 00:20:12 here click download it will get downloaded like this as you can see.

• 00:20:17 Okay, once downloads has been completed, go to your downloads folder, select the files,

• 00:20:24 click cut, and then go back to the installation folder of the web ui and in here go to models

• 00:20:34 folder and in here go to stable diffusion folder and this is the place where you are

• 00:20:42 going to paste, put your full models that you have downloaded.

• 00:20:46 I am explaining everything in the other video that I have mentioned and then go back to

• 00:20:51 web user dot file.

• 00:20:55 Okay, I'm closing the other command windows and this time it will be much faster because

• 00:21:02 the initial step has been completed.

• 00:21:04 You see it is launching the web ui with the arguments minus minus xformers and minus minus

• 00:21:10 no half.

• 00:21:13 Okay, okay so it is ready, but you see it didn't open me any pages.

• 00:21:20 So you need to type this into your browser or you can copy and paste it.

• 00:21:25 I am selecting it with my mouse right click actually right click not working.

• 00:21:29 You need to select it with your mouse, then control-c it will copy it and then you need

• 00:21:35 to paste it into your browser and it will get opened.

• 00:21:40 Currently it is running.

• 00:21:42 Let's generate a new image with the same prompt.

• 00:21:46 I will post this prompt into the comments section of the video.

• 00:21:51 I am selecting DPM++ SDE Karras this is the sampling method and this is one of the good

• 00:22:01 ones that I have found.

• 00:22:02 Then click, generate and and you you are seeing that it is generating each step.

• 00:22:10 You will see the progress here and also here.

• 00:22:12 Okay, the image has been generated, but you see this is pretty different than the first

• 00:22:19 example.

• 00:22:20 Why?

• 00:22:21 Because currently it is using anything v3 model.

• 00:22:25 However, in the previous example, we were using version 1.5 which was the official version.

• 00:22:32 If I change back model to that version and generate another image, let's see what we

• 00:22:38 are going to get and this is what we get with the official 1.5 version model.

• 00:22:45 Okay, this is all for today.

• 00:22:48 Please subscribe our channel and follow our videos because I am planning to make a full

• 00:22:55 course regarding stable diffusion with explaining pretty much everything about it.

• 00:23:01 You can see it in our playlist in the stable diffusion.

• 00:23:06 I will put them with an order.

• 00:23:09 I am generating videos both for Google Colab and that will run on your computer.

• 00:23:16 To make it run on your computer, you need a decent graphic card.

• 00:23:20 You see, even my RTX 3060 is struggling a little bit.

• 00:23:29 So therefore, if you don't have a strong GPU you can use Google Colab or other services,

• 00:23:35 but they are a little bit harder to use than installing on your computer.

• 00:23:41 Please also consider liking sharing this video, subscribing, and also supporting us on patreon.

• 00:23:50 This will enable us to continue produce more high quality videos in the future.

• 00:23:55 Currently, we still have zero patrons, but I am hoping that you will be the first one.

• 00:24:01 Hopefully see you in another video.