I'm creating a LoRa to diversify the faces I get. It's a personal project, and I haven't decided to make it professional yet. I'm just a beginner, and I'd appreciate if anyone is enthusiastic enough to help me on this 😍 I'd share the benefits 💕
All the subjects are female because I couldn't find any website that rates male faces 😁 Currently I'm using babepedia.com 's ratings.
I'm genuinely sorry for repetitive upload of V0.3. I'm getting used to how Civit website works.
Version 0.1 "Diversification/Nationalisation": Just a test run with one person.
Version 0.2 "Natiodiversification": Added three more girls, all of Persian descent.
Version 0.3 "Natiodiversification": Adjusted the over-cooking in V0.2
V0.4: I will go for the Middle East and other parts of the world.
Description
Adjusted the over-cooking in V0.2
FAQ
Comments (4)
Wow is this "mix-of" a default query of SD? Or someone added it in training at some point?
@KoiArt default. hyphens are weights used in sd and can be mixed wildly
@muetze Mhm. Now I looked it up and found this:
https://civitai.com/articles/3875/hypens-dashes-and-implicit-weights
Didn't know dash and hyphen differ here 😮 Thank you for the info 😊🌼



