Automate your Stable Diffusion!

Key Takeaways

do you wish you could automate your stable fusion and let the machine do all your tasks well you can so let's check out this next part where I show you how and by the way don't try to spell part backwards it's a trap I will show you how to set up tasks like these automate them and let the machine do the rest so art Venture X has something called the agent scheduler I'm gonna link this in the description below so just go here press the code here press copy here and then when you go into your automatic 1111 you can find extensions here and this is not available in the list here in automatic 1111 yet however it is in vladder Fusion or SD next as they're calling it now so if you're using Vlad you can just load it from the list however we're just going to install from URLs take the copy paste the link put that in there and press install now when you install it and reload it you're gonna see agent scheduler up here and also the enqueue button here so instead of generating you can push this to send it to Q now if you go into the settings here you can find agent schedule down here as a default we have disabled queue Auto processing so when you're curing something or automating it doesn't start if you want the generations to start as soon as you queue them you don't have to uncheck this button we're going to leave this as is for now now let's go back to into our text to image we're currently using the deliberate model we're going to load our styles to default negative and the digital oil painting these can be found in the description below free as usual let's put in cat in a hat let me just wait that up a little bit with Ctrl Arrow up and we're using four images and let's just cue that now let's change into penguin hat and we're queuing that up and let's change the model here to Let's Take protogen and let's put in sci-fi penguin in a hat and let's thank you now we have put in three cues and if you go into agent scheduler here you can see what you have we have three prompts a cat and a hat the Penguin and a hat and a Sci-Fi penguin and a hat with different checkpoints we have deliberate on two of them and the protein here now you can change more settings like the sampler steps scale and size but for now we're gonna be happy with this now you can start in one by one here or you can just press resume here to run all of them this is a great way for you to start like if you want to render overnight you can start a process of doing lots of different things so now you can see it starts processing here now we're getting some cats we got some women as well this seems to be a penguin whenever this is finished you can find it in task history here so you can see we have Katana hats now if this is a woman obviously but we got our four Images here um all the settings now you can't see them because of my face but uh it's not that important you have the generation info here similar as the PNG info that way you can run something for example overnight or just test a lot of different settings instead of using like the XYZ plot and here we have our penguin and a hat and our images and here we have our last one are supposed to be sci-fi penguin and hat that didn't really turn into sci-fi but you have to prompt a little bit more for that probably so now the cool part about this is you can actually re-queue these so if you have something you like maybe you like this prompt here the Penguin and a hat and all the settings you can save them you can bookmark it here to remember it for later and you can just okay I like this one this I ran 10 of these overnight this was the best one let's bookmark it so I remember and then I can just recue it and it's going to be right back here in the queue and we are now generating it again now if you don't want these to auto run from the Wii queue you're gonna have to press the little pause button here now you won't see your Generations in the text to image here so make sure sure to check that you're in the agent scheduler when you're pressing the enqueue button so everything is going to be in here now if you found something that you like now I have four images of women in the jungle here so semi-realistic portraits you can like we talked about run it again we could also send it back into text to image imaged image in paint to extras and that's done with the buttons that you can't see behind me so I'm just gonna show you here so we have text image here if we send that right back we're getting all the settings that we used so this is a great way to save the settings so it even saved the Samplers the steps with the height seed everything so it's a super smooth way to just get your settings back and I think this feature is one of the best if not the best feature of this extension this is a really cool and effective way to automate your workflow so I hope you like this little tidbit of information check back in for more tips on generative Ai and stable Fusion as always have a good one see you in the next video