Recommended CFG scale : 7/6.5
Recommended LoRa weight : 0.9/0.95
>To use my LoRa,
Use kohya-ss/sd-webui-additional-networks (github.com) in auto1111 to load the LoRA model. You can combine it with other LoRA for blended poses.
Trigger Words : "girl" ,"woman"
Description
FAQ
Comments (11)
Looks promising, I will check it out at my spare time.
thank you <3
Tried it. It works so well! I am trying to recreate something similar with my face, but every time I train a model it looks nothing like me. It's not even close!
Could you tell me how many training images you used, and what training settings you used for this and your other models?
Do you use regularization images or do you just caption the photos?
Thank you man, this is great, can't believe Lora from 10 months ago is that good, incredible work
Dogu you got way better at loras since making this one :O it's not bad! it's just a bit cartoony-animey, would you consider trying a 2.0? your recent work has built on this a lot