Patreon
Get early access to build and test build, be able to try all epochs and test them by yourself on Patreon or contact me for support on Discord
Recommended settings
Make sure you are aware on the usage instructions of LORA
Sampler: DPM++ SDE Karras (Recommended for best quality, you may try other samplers)
CFG Scale : 5 to 10
LORA weight for txt2img: reccomended 0.2-0.5.
Insutrctions
My advice is to start with posted images prompts then change them, in any case you will find different keywords for each model (wich should be used with a weight of 1.2-1.3).
Also in version description on the right you will find more details about each hairstyle model.
Enjoy!
If you'd like to support my efforts and have access to early and bleeding edge builds, please think about joining my Patreon: https://www.patreon.com/betterCuteAsians
Description
This hairstyle model doesn't always work in frontal view, you need a lucky seed.
But gives good result for side and back view. Because of the training images, mostly from side and back cause of the hairstyle type.
For frontal view you might want to raise the weight of model/keyword but you might get artifacts at some point. For side view you might wont to lower it.
Works also with anime, look at examples.
FAQ
Comments (9)
Thanks for posting these, any chance you can make one for a long braid?
It might be too hard to find proper images, making bridal was already really hard. If you can find the images for me I ll make it otherwise I m not sure I can do it.
@antonio_riolo2610 Thanks! I'll see what I can do. What are your criteria for image selection?
atleast 512 width and 512 height, any size higher then this, any shape. atleast 3 images from behind, atleast 5 images from side and atleast 10 images from front (frontal are the hardest to find for hairstyles because you have to pick the one where the hairstyle it's recognizable, and in frontal view sometime it's very hard to find an image that points out the particularity of that hairstyle), for a total minimum of 20 (max 25 it's fine). it would be nice to have most of the images with a background, anything that is not a monochrome background or the model will learn also the no background concept making very hard to generate good images.
I'd say all realistic images but if you can't find enough realistic, a few anime images can also work, but mostly realistic.
Good Luck!
@antonio_riolo2610 Thanks! I'll start looking for that!
@CallMeMaybe long braid hairstyle is in training, just i'm not sure if the images i found were what you was looking for. I'll pubblish it later and i hope you will like the result :)
@CallMeMaybe Here https://civitai.com/models/76937?modelVersionId=87048.
I hope it's what you was looking for. I'm really proud of the result.
Enjoy!
@antonio_riolo2610 WOW! I was going to try to get to it this weekend, but you just nailed it! Thank you so much <3
@CallMeMaybe Thanks!!! It wasn't easy! You might leave a review if you like it :)
Details
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.