Follow me on X/Twitter for exclusive content: https://x.com/Citron_Legacy 😁
Check out all my Style Loras: https://civarchive.com/collections/10766871
I intentionally used "sketchy" art so that this Lora generates not rough art not professional tier art.
Description
FAQ
Comments (13)
caught wondering how this might actually help people who's learning to draw... one thing for sure though no one's gonna fault these drawings when they're missing a finger or two XD
I was thinking the exact same thing about bad hands when I made this lora! lol
This is certainly not a professional grade tutorial on drawing, but it might spark some ideas.
I think the mistakes in this art makes it a bit more approach able for someone trying to learn because its less intimidating.
I was thinking if someone needs to help coming up with basic things to sketch they can use this to generate "drawing prompts"
Unlike professional colored art this makes it extremely easy to see where lines should be drawn and also provides guidance on where to apply line weight. These images actually do a really good job showing how thick vs thin lines make an image look nice.
@CitronLegacy very true! if anything it at also really helps out people trying to learn coloring without having to make sketches from scratch haha
I mean, yeah though honestly any image can. Studying what you see and replicating it and working backwards to do so has always been one of the foremost ways of practice.
That being said, if you are not learning say anatomy the proper way, AI makes mistakes and you may not know they are even mistakes you are copying; if you never learned it from an accurate source.
Still, I think in moderation Ai could absolutely help, and may also keep things more interesting and keep you more engaged.
@pihlawrkr738 I agree that if you aren't learning anatomy the right way then you are shooting yourself in the foot. My foot has been shot repeatedly lol
I have this hope that someday I can make a lora for each step of the drawing process.
For example the very first step would be the anatomy "primitive shapes" (this is what its called it in 3D modeling I'm not sure what the correct term is for 2D art?) -- if you prompt 1girl it would give you circles, spheres, ovals, triangles of body proportions rather than a finished picture.
@CitronLegacy There is a model that adds topography to images, you can use that to get a general understanding; and then further break it down into your own forms.
@pihlawrkr738 That sounds familiar but when I search for it I cant find it. Could you please link it?
I was considering making something like this so if it already exists that would be nice.
may you do the same for SDXL too?
Thanks for the reminder!
I made the SDXL version awhile ago but didn't post it until today.
@CitronLegacy Issue: Anime_Sketch_SDXL.safetensors: AssertionError, Anime_Sketch_SDXL.safetensors is not a safetensors file, getting this error
@krigeta Thats weird. Are there any other errors?
I was able to use this file to generate images via the CivitAI image generator.
For example, all of the images in this post were generated on CivitAI https://civitai.com/posts/979701
@CitronLegacy its working with that exception