How to install and use Stable Diffusion in 6 minutes

why pay for Mid Journey when you can generate images with artificial intelligence right on your computer and best of all for free without limits and with total privacy today on AI I'm going to show you how to install stable diffusion on your own computer and in less than 10 minutes generate fantastic images using this free AI let's go stability AI launched stable diffusion in August 2022 as an artificial intelligence model capable of generating images from from text in October version 1.5 was released which to this day is the most widely used because it is lightweight and comes with a free license soon there was a legion of enthusiasts who retrained the original model to create new refined models known as checkpoints these refined models are hosted on civit AI and in today's video we're going to use cyber realistic one of the best checkpoints for stable diffusion 1.5 but before we start I need I need to tell you that stable diffusion is an AI model that needs an Nvidia GPU to work so if you don't have an Nvidia GPU unfortunately it won't work on Windows now let's get started create an account on civit AI and download the latest version of cyber realistic by clicking on this little blue button and choosing the first option also download the negative embedding for the model which will be used to generate even better images while the files are being downloaded let's install the stable diffusion web UI stable diffusion is not an application that can run by itself instead it's just an AI model so in order to use stable diffusion you need a program to manage this model there are several options but today we're going to use the stable diffusion web UI also known as automatic 1111 this is the most widely used web interface for stable diffusion and you can find lots of tutorials on how to use it on this channel I've made over 60 videos about it so if you're enjoying this video help me make this channel grow even more by clicking the like button and leaving your questions in the comments you help me and I'll help you generate your images on your computer back to the video now let's go to automatics GitHub page and in the installation instructions it says that we first need to install the Python programming language and the Git Version manager so we go to the python page and download version 3.10.7 the installation process is simple but don't forget to check this box before you start git is even easier go to the git page and download the 64-bit version for Windows then just install it using all the default options now I'm going to create a folder called programs and add this folder to the Explorer shortcuts so it's easier to find where I've installed stable defusion now I go into this folder and type Powershell at the top a terminal window will open I go back to the installation instructions and copy this command I paste the command into the terminal and press enter now automatic is being downloaded and if I look in Explorer I can see that it's already inside the programs folder I just created now I just have to copy cyber realistic to that folder where the templates live I also have to copy this negative embedding file to this other folder now I'm going to edit automatics startup script to add X forers this will make stable diffusion 30% faster I'm also going to add autol launch so that every time I start stable diffusion the browser opens automatically now I'll create a shortcut to this file and that's it just doubleclick it to start stable diffusion this first run will take a while to open because automatic will install all the libraries needed for stable diffusion to work you can go for a coffee and come back in a few minutes that's it a few minutes later stable diffusion is up and running let's generate our first image by typing in a very simple prompt and clicking this button great but it could be better I'm going to add more tokens to my prompt to describe what this woman looks like where she is and a few more tokens that improve the quality I'll also add the name of the negative embedding down here to improve the quality of the image there it is perfect now I'm going to simplify my prompt using a feature known as Styles as you've seen only part of my prompt describes the actual image most of it is just a bunch of tokens I've added to enhance the quality so I'm going to click here to create a new style called realistic and I'll write here all the tokens that describe the style and not the image itself then just save and choose the realistic style from this list every image generated from now on will include the realistic style tokens without me having to write everything manually if you want you can change the image size to create square or portrait formats but if you just want to generate higher resolution images instead of changing the size manually turn on the highres fix option leave the seed fixed to always generate the same image and change the upscale rate to 1.5 to get a a 50% larger image with more details here's how it looks you can also make small adjustments to The Prompt now in this case I'm changing the woman from a blonde to a red head look at the final result and if you want to see all the images generated go to this folder within the automatic installation all the images will be here separated by the day they were generated and that's it in just a few minutes we've installed stable diffusion on our own computer we've also learned how to generate our first highquality image and all for free and without limitations now if you want to make your stable diffusion even better take a look at this card see you in the next [Music] video

Channel: Today on AI

Views: 1,995

Keywords: stable diffusion tutorial, stable diffusion automatic1111, ai images free, how to generate images with ai, ai art

Id: NlB-RlmlkA8

Length: 5min 49sec (349 seconds)

Published: Fri Sep 29 2023