BASIC - All you need to know about Stable Diffusion (Mac & Windows)

hello everyone with this video I would like to give you an overview of stable diffusion of all of the interfaces we can use for running it their requirements and this is mainly for Mac users and also alternative solutions for running it not just locally B also remotely stable diffusion is a AI model to generate images and so videos as well from a text description so unless we are very good with coding how can we interact with this platform there are some interfaces we can use the most famous is automatic 1111 but there are also others like invoke AI for example or compi automotic 1111 is my favorite is Super User friendly and you have a lot of tools for generating different images customize them in painting a painting you can generate videos but what's best is that you can also customize the interface it's continuously up to dat meaning that there are people developing it and a lot of people generating extensions for it so a lot of things coming up almost every day it's very difficult to keep up with that but it's honestly is very cool another one is invoke is very very similar to automatic 1111 but you have less tools for customizing your image the interface is user friendly as well so you don't really need experience for running it confi is the last one I want to talk about is super cool as well it's less user friendly I would say but just initially just initially because you need to know how stable diffusion works all of these three interfaces can be run locally or remotely if you want to run them locally you need to have some software requirements very important we'll talk about this in a minute but you also need to deal with coding because in order to install them on your computer you need to open your terminal and you need to write some codes on your terminal in order to run it it it's usually quite easy because you just need to run requirements and then you need to install mainly for Windows but sometimes can be a little bit more difficult because you can have different requirements on depending on your computer and you can end up having some issues when running it now talking about requirements stable diffusion requires at least 6 Giga of vram although 8 GB are recommended if you want a better performance and mainly if you are generating B images or long videos Nvidia gpus meet this requirement and actually exceed them and this is what makes the difference between having a Windows computer and a Mac AOS Mac system and mainly the newer models with M1 and M2 chips integrator do not support Nvidia gpus and this lack of support for NVIDIA GPU can be a limitation for Mac users so I'm not saying that I cannot run stable diffusion I actually have a M1 and I can run it but it's extremely low performing it's very slow it takes ages before generating an image it's probably better if you have a more powerful Mark but still it's not the best is not as good as having a Windows however there are some workarounds in order to run stable diffusion from any computer even the least powerful so let me give you these two alter natives the first one is to use cloudbased platforms that provide user with virtual environment and these are collab rumpot paper space And there are many others I'm sure but I'm going to mention just the most famous One Running automatic 1111 with Google collab used to be free right and this was I think up to March 20123 or something like that it was amazing so I was running actually like with my Mac collab stable diffusion on collab for free however since then we need to upgrade to collab Pro to be able to run stable diffusion automatic 1111 on Google collab the monthly fee is around $10 or something like that and collab Pro offers better gpus and 100 computer units per month instead of automatic 1111 you can also run the invoke Ki interface on Google collab and also G fui and I can tell you that these two are free to run so you don't have to pay anything however with Google collab there are some issues sometimes because I had cases where my connection was breaking all of the sudden so I was losing all the things I was doing and I noticed that this happens quite often with Google collab actually another downside is that uh you every time you want to use collab you need to rerun the full code if you prefer not paying a monthly subscription but you want to pay as you go you can try ramp pod instead ramp pod offers GPU power virtual machines and it's it's actually quite good and as for a Google collab you have templates already for running stable diffusion so you just need to run the code so which one would you choose well it depends on what you need and how much you're going to use stable diffusion I find Rod a little bit more um reliable usually then the second option if you don't like coding and you don't want to deal with this connection issues you can go for cloud-based platforms instead in this case you don't have to do anything with coding you just need to sign up log in and then press the Run button for running stable diffusion there are different platforms available so you have the most famous run diffusion diffusion Hub and think diffusion as well let's have a quick look at the pricing together this is run diffusion and you can see that they have three different plans you have the pay as you go which is 050 per hour and then you have the professional plan where you pay a monthly service and then business and Enterprise if we scroll down you can see the different features you can uh you can have and then you can see that you can use stable diffusion 1.5 stable diffusion Excel and depending on what you need depending on what you want you can pay different amounts and get different uh speed or different different resolution as well this is the fusion Hub also in this case you have uh pay as you go and then in this case you don't just have automatic 1111 but you also have Coya and confi once you launch the platform is exactly the same as automatic 111 you run locally with the difference that you're not running in locally but remotely and so you don't have to codee anything you don't have connection ises and I know that they were working on a subscription monthly subscription uh but I see here that you can choose different plants you have basic premium ultimate and you can insert a referral code if you want actually if you want to try this I have a referral code you can use it's down in the description the last one I want to show you is think diffusion this is their website and as you can see they don't just have automatic 1111 and confu ey they also have fogus in and Coya so also in this case you can choose different you know GPU requirements and vram so this is the turbo one it's the one with highest vram highest GPU but you can also go for a cheaper option but you can just run stable diffusion 1.5 while with this other one you can run also stable diffusion Excel and control net which obviously makes the difference when you're working on something more professional I think this the same uh it's for all of the others yeah you have different options for for all of them so that's it for today I just wanted to you know explain why Mac User see this very low performance when running stable diffusion Lally and that's why I wanted to give you alternatives for running it and being able to generate images faster and um more efficiently so I hope this was useful thank you very much for watching and see you in the next video bye but

Channel: Laura Carnevali

Views: 1,091

Keywords: midjourney, stable diffusion, dream studio, stable diffusion v2, diffusion, ai art, diffusion model, generative ai, generative art, dalle 2, dalle, openai, openai dalle, stability ai, ai artist, imagen, nft, install stable diffusion, install stable diffusion on mac, install stable diffusion apple silicon, apple silicon, m1, stable diffusion on m1, stable diffusion with python, stable diffusion hugging face, stable diffusion github, stable diffusion v1.5, stable diffusion tutorial

Id: GZEIUpDgVOg

Length: 8min 30sec (510 seconds)

Published: Thu Feb 01 2024