FREE 2023 Stable Diffusion PC INSTALLATION! AI Art For BEGINNERS!

Key Takeaways

and as promised in my last video of 2022 I said that I'm gonna be updating my older videos to 2023 and what better way to start these updated videos than with the installation of stable diffusion and since I haven't installed it already on my new computer not only I will be installing stable diffusion on my new computer live but I will also show you how to do it too so hope you see this video is only for the beginners this is only for people who don't have simple diffusion installed already on their computers so if you're already part of my community in this renew is probably not for you alright then let's begin oh and also quick plug I know that I said in my previous video on Friday that my first video on my new second channel the maker layer was live but unfortunately I had to wait way longer than I thought for YouTube to verify my identity don't ask it's just their new policy so if you're interested in seeing Oh I got monetized on YouTube in less than two weeks the link for this video will be in the description down below that being said let's begin alright so before we start the installation of stable diffusion you need a few things first make sure that you are using Windows and if you have an Nvidia card otherwise this installation will not work for you now there is a way to install stable diffusion on Apple computers and on Linux but since I don't have these machines I cannot show you how to do it but I will leave a link in the description down below for all of you apple and Linux folks also if you have windows installed but you have an AMD GPU this install will probably not for you either but again I will leave in the description down below and this solution guide for all of you AMD users so make sure that you have a Windows install a good Nvidia card with at least 4 gigabytes of vram although you can run it if you have at least 2 gigabytes but 4 gigabytes is definitely the recommended amount now before we can install stable diffusion we need to install two things Python and git for python make sure that you're downloading the 3.10.6 version because stable diffusion might not work if you use a newer version and again the link for this page will be in the description down below so if you click on it you can arrive on this page you're just gonna scroll down and then you're gonna click on Windows installer 64-bit then you're gonna do the same thing forget you're gonna arrive on this page and then you're gonna click on 64-bit git for Windows setup and once you have these two files let's start by installing python so you double click on the python XD file and here don't forget to click add python 3.10 to path and then click here to install now and this will start installing python on your computer and simple as that python was successfully installed onto our computer now that python is installed you're gonna double click on the git.exe file to start the installation here in the setup menu you're gonna click next You're Gonna Keep clicking next now for the default editor personally I'm going to choose notepad plus plus and then keep clicking next until git is installed and then click on finish and now that we have both Python and git installed you can delete these two files because we don't need them anymore so now you're gonna create a brand new folder where you're gonna put stable diffusion and I highly recommend creating a folder without any spaces because this can be a source of issue later on so for this you're gonna right click click on new and then you're gonna name your folder anything you want in my case I will call it super SD I think my subscribers will know what it means I think gonna click on the folder URL type CMD and then press enter and this will bring the command prompt window and inside the common prompt window you're gonna copy and paste that command that you will find in this description down below and then you're gonna press enter and you see that this Command right now has created a new folder called table the fusion web UI and if you go inside that folder you will see a bunch of files so the thing that you see right here is basically a copy of the automatic 1111 repository on GitHub this is the web UI that you're gonna be using from now on but all the files that you see are just for running the web UI to be able to use stable diffusion you need to download stable diffusion models and this is where if you are a complete beginner you're gonna have a big problem because there is a lot of models to choose from even if you only count the official ones there is the stable diffusion 1.4 there is the same with the fusion 1.5 this special 1.5 in painting model the 2.0512 model the 2.0768 model the 2.0 depth model the 2.1 512 and the 2.1 768 yeah there is a lot of models to choose from and again as I said these are only the official models released by stability Ai and Runway ML and we're not even talking about the hundreds of different models created by the community which are way way better than the official models so then which one should you choose which one should you download well you can download either all of them or neither and you can just download a few models made by the community for example the protogen 2.2 which I've made a video about on Friday the link for that video will be in the description down below or if you want to have the official models but don't want to download all of them here is my advice you can download the 1.4 1.5 and the special stable diffusion in painting model these are my recommendation these are the best models for you to learn how to use it these are the best models for complete beginners because they're easy to use and there's plenty of examples available online for you to take inspiration from now you might think well when it comes to 2.0 2.0 is higher than 1.5 right why can't we just download this instead now I'm not gonna go into that whole drama situation again if you want to know what happened I will again leave a link in the description down below for that video the 2.0 models and the 2.1 models are a little bit worse than the 1.5 because these models were trained differently and because of that there are not as easy to use and also produce results that are not as varied compared to the 1.5 for example so if you are a complete beginner and you just want to have fun download the 1.4 1.5 ending in painting model and if you also watch my previous video on Friday about protogen you can also download the protogen V 2.2 model that will make better images than all of these models combine also if one day you decide to download the 2.1 or the 2.0 models I highly recommend just skipping over the 2.0 models and just downloading the 2.1 since they are better than the 2.0 now in this example I will only download the 1.4 and the 2.1 and I do that because there is a small difference if you want to install the 2.1 model then if you want to install the 1.4 you can simply click here in files and versions and then click on sdv14.capt this is a 4 gigabyte file and that is what we call a model or a checkpoint and then click on download and I'm also going to do the same with the 2.1 and I highly recommend if you want to use the 2.1 then you choose the 768 version because it creates way better images at higher resolution so again just click on files and versions and then you're gonna click on EMA prude dot safe tensors the non-em8 versions are only for fine tuning in training so click here and then click on download and then again for the 2.1 version you're gonna click the link in the description down below and you could have to download an additional file that ends in dot yaml this file basically tell the UI to correctly load the 2.1 model so yeah as I said the 2.1 requires a little bit more work so if you choose the 2.1 768 version you're gonna have to download this little file right here and in the end you should have three files you stable diffusion 1.4 model dot ckpt the 2.1 save tensor file and a small yaml file and now it's very easy all you have to do is just select these files Ctrl X to cut them in your stable diffusion web UI folder you're gonna go into models table the Fusion you're gonna paste these models right here now since this is an EMA model you're just gonna rename the yaml file from non-ema pruned to Ema pruned.yaml and now we can go back and double click on the web UI user.bot file but one thing that you can do to avoid any issue is to right click edit with notepad and in your start menu you're gonna type Python and then click on open file location then right click again and open the file location again and you should see a similar folder this is the folder where you install the python and what you're gonna do is right click on python.exe and then click on copy as path then you're gonna go back to your notepad and after set python equals you're gonna paste the path right here and then click on Save and this little procedure will actually solve a lot of issues linked to improper installation of python so just in case just do this this could solve a lot of potential issues so after this is done all you have to do is just double click on the web you my user.bat and this will start the final installation of stable diffusion I just need to say that it will take a lot of time for some people it could take up to 30 to 40 minutes yes I'm not kidding so if you think that the installation is stuck you really just need to wait don't be patient let the installation finish just you know go outside and touch some grass so yeah just be patient and everything will be fine and there you go stable diffusion is finally ready so for me it took around 15 minutes and I have a pretty beefy computer with a very fast internet connection so I know that for some of you this installation might take a lot of time but once this is done you guys see here a local URL and if you select it then Ctrl C to copy it and then you're gonna paste that address into your browser and then press enter and there you go now you have stable diffusion running on your computer and all you have to do is just go to your stable diffusion checkpoint select the model that you want then you're gonna write your prompt choose your sampler the number of steps CG scale and then click on generate and this is the final image I almost forgot how bad the 1.4 looked definitely the models that we have right now work way way better and if you want to change the model all you have to do is just click on stable diffusion checkpoint and then select the model that you want to use and in my case I want to use the 2.1 768 model and since this is the 768 model I'm gonna increase the width and the height and then click on generate and this is the final result now I know for the 2.0 and the 2.1 model you're supposed to use a lot of negative prompts to make your final images look better but since this is just a test I'm not gonna bother now if you have trouble using the 2.1 model one thing that you can do is right click on the web UI user bot file edit with notepad and here after set command line arjs you're gonna type Dash Dash xformers and then click save and then you can relaunch stable diffusion as you can see now it will start installing xformers which is a library that will allow you to generate faster images with stable diffusion and now stable diffusion should be running for you perfectly but if it isn't I will leave a link in the description down below for an FAQ that two amazing people from my Discord server wrote for anyone having trouble with stable diffusion it is fairly complete and it should answer most of your installation problems so a big thank you to spacegender and Camille Lorex for writing this amazing FAQ and if you still cannot find your issue you can join my Discord server and our community could potentially help you and as always all the links will be in the description down below also when it comes to updating stable diffusion you have two solutions the manual way and the automatic way for the manual way you're gonna come here and click on your photo URL type CMD and press enter and here in the command prompt window you're gonna type git pull and then press enter and just like that it will download the latest version of stable diffusion web UI and for the automatic method you're gonna right click on web UI user web file click on edit with Notepad and here just above call webui.bat you're gonna type git pull and then save the file and now each time that you want to launch stable diffusion it will update stable diffusion web UI automatically and there we have it folks now you should have stable diffusion installed on your computer for absolutely free and you are finally ready to make some amazing AI art in only a few seconds so get out there and make some cool AI art and there you go thank you guys so much for watching don't forget to subscribe and smash the like button for the YouTube algorithm and I'll see you guys next time bye bye