Stable Diffusion Basics - How to execute many prompts at once AND specify settings for each one

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
hello I'm silicon thamaturgy and welcome to stable diffusion Basics short tutorials covering single and hopefully less obvious features in the automatic 1111 web GUI today I'm going to show you how to use prompts from file slash text box this allows you to execute a preset list of prompts as a batch instead of individual prompts the main downside of this to me is that you can't use it to generate X Y plots with which as you've seen from my previous videos I use at any conceivable opportunity I've mainly used this feature to test embeddings I've made on a variety of prompts without having to enter each prompt individually on the other hand you might want to generate a bunch of pictures of Emma Watson or Christina Hendricks I mean theoretically a couple of notes on what you can and can't use with this I've tested this using one and both Styles up in the upper right which are selected by the drop down and they do work also you can use the high-res fix on this though I will say if you have a lot of prompts to run through it will take a long time to finish so maybe better just pick the ones you like best after generation run an image to image on them now let's try it out first assemble a list of prompts that you want to test I usually do this in a spreadsheet instead of a text file because I'm not a masochist I got most of these prompts from Lexica so thanks to it or put them there I replace the subject of all these prompts with three hash signs this is arbitrary and simply so it can quickly and efficiently replace that with whatever subject I want if you're willing to do some text formatting you can also change all the variables per generation for each job however this does not apply to the variables for high-res fix instead of having your prompt as just a single block you'll need to identify each variable with a Double Dash followed by the variable name then the value of the variable variables that are text require quotation marks around them while variables that are numbers do not when a variable is not specified for a particular prompt it will use the settings from the interface instead here are the results from the original prompt a photo of a tree and the four modification shown in the last image well that's enough nerdy stuff let's get this show on the road next we're going to go into the list of prompts and do a find and replace to ensure they're a subject whether embedding or Emma Watson isn't all the prompts from here you have two choices you can either copy the text directly into the text box in the stable diffusion web GUI interface or you can say this as a text file and then drag and drop that into the stable diffusion web UI interface after that all that's left is to select the settings you want and click the generate button thank you for watching if you found this video helpful please like And subscribe if you have ideas or requests for future videos don't be afraid to leave a comment
Info
Channel: SiliconThaumaturgy
Views: 7,995
Rating: undefined out of 5
Keywords:
Id: 9PhZ2r1gnt8
Channel Id: undefined
Length: 2min 59sec (179 seconds)
Published: Thu Feb 09 2023
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.