Easy run SDXL in A1111!

hello everyone today I would like to show you how to use stable division Excel version 1 within automatic 1111. I haven't been able to install it on Mac yet because I'm getting a lot of issues at the moment so I'm gonna go through windows so hopefully will be helpful actually on Windows it's super easy so if you have never installed it you just need to go into the main guitar page of automatic 1111 which is this one and then you copy this code in here you open your command common prompt then you check where you are in the directory you dive in the folder where you would like to install your automatic 1111 in my case I want to do that on my documents let's do that so I will change directory and type document and then I will get clone the link I've just copied from the Gita page so here it's giving me a warning press yes then it's going to ask me to enter my password for the laptop I'm gonna enter it so when you type the password inside your command line you cannot see it so if you're typing but you cannot see anything coming out and no worries something is happening so you just insert your password and press enter and then we'll get all of the data from the GitHub page super quick so now if I go into my documents I'll have this new stable diffusion web UI folder with all of the different files from GitHub and this is the first step now if you have already stable diffusion instead what you have to do I have this stable diffusion ybi on my desktop okay so I'll open my command line again I'll see where I am I'll dive into desktop and then I'll dive into stable diffusion web UI so change directory stable diffusion web UI let's see what's inside okay cool and then here we just need to type git pull so this is going to update my folder with the new updates from GitHub is going to ask you some permission you need to type yes and then you will need to enter your password again of the computer also in this case if you don't see anything while you type don't worry something is happening in reality so just tap your password and press enter and then it's going to pull all of the information from GitHub and it's going to update your local folder in my case I did it already so it's already up to date okay and the first step is done now we need to download the stable diffusion version 1 models and drag and drop them or copy and paste them inside our new folder we go into again phase in the stability AI page and then we can go into stable diffusion Excel base 1 and Excel refiner let's go into this one first we're going to file some version and then we download the save 10 Source this one here and we do the same for the refiner so let's go back file and versions and this one if you want you can also download the variation of the encoder foreign ER model by the way but yeah it's this one here so you can download it and then you can add it into variational out encoder folder now once you have that loaded you will have them off obviously in the download folder what you have to do I'm gonna show you what I did you go into your stable diffusion web UI folder your local one obviously you go into models and then you have stable diffusion and then here is where you need to add your stable diffusion Excel base and stable division Excel refiner here they are one and two for the variational encoder instead you need to go into variational audio encoder folder and you will add I have this one as the Excel variational of the encoder in this folder here once we have done this we can initialize our automatic 11 11. so we'll double click on the web UI user but okay all good great and we need to open this link here and now we have stable diffusion so the next step is to change the checkpoint to have the base in here so we'll look for the stable division Excel base model okay you will change the stable diffusion variation allow the encoder to be automatic or known or you can also use the variation allowed encoder you downloaded if you don't have this you need to go into settings and then you scroll down I usually go into show all pages because I never remember where to go and then in here where you have quick setting list you can add this stable diffusion variational encoder you can add different options and all the options you will add in here they will show up on top of your main page so once you've done that you need to apply settings and reload UI cool so once you've done it let's go back to the text to image we type something like I don't know a woman and a dog running across a field and then we can add I'm gonna steal some prompt from realistic vision we don't need to create the a very detailed prompt description with the new stable diffusion Excel just for you to be aware for getting realistic pictures but surely helps having something in the negative prompt at least so I'm going to remove bad dream and unrealistic dream and then maybe here I cannot realistic detailed 8K I'm going to change this to 30 sampling steps because it seems to be better I'm going to use DPM 2 or maybe I can use DPM plus plus SD cards I'm going to change the resolution to the default one for this model which is higher so it's 1024 I'm Gonna Leave the seed as it is and then I'm going to press generate so in this case we are running only the base Excel model we are not running the refinement stage we'll do this in another step of afterwards and this is the picture we got with the base model now this has to be the input for the refiner right so we'll send this picture to image to image we'll change the checkpoint with the refiner which is this one okay then we scroll down and I'm going to use the same sampling method 30 steps I'm happy with 1024 I'm happy with the CFG scale I'm going to change the noisy strand so the refiner works exactly the same as the image to image right so the higher the nosy strength the more different will be the the final output compared to the image generated by the base model the lower the most similar the output will be to the image generated by the base model so if we use adenoid strand equal to zero we won't have any change we should use something between 0.10.3 I would say so in this case I'm going to use 0.20 so that we will have a very similar image to the one we just generated but with more detailed right like nicer so let's press on generate and let's see what happened and here you go it definitely looks better obviously there's still something that we should change like the hands for example that will look great but I think that overall it looks okay that's it for it today this was a very quick one I just wanted to show you how to use stable diffusion Excel version one within automatic 11 11 and I think it looks great it works well and this is Windows I'm using an Nvidia 3090 GPU so with Mac it's slightly difficult so I hope this was useful and see you at the next video bye

Channel: Laura Carnevali

Views: 4,764

Keywords: stable diffusion, diffusion, ai art, diffusion model, generative ai, generative art, stability ai, ai artist, imagen, nft, install stable diffusion, install stable diffusion on mac, install stable diffusion apple silicon, apple silicon, stable diffusion on m1, stable diffusion with python, stable diffusion hugging face, stable diffusion github, stable diffusion v1.5, stable diffusion tutorial, stable diffusion xl, refiner, refinement stage, new sd, sdxl 1.0, sdxl, sdxl base, v1

Id: 2UMDEhGEjos

Length: 9min 14sec (554 seconds)

Published: Tue Aug 01 2023