why?
To create closeups to medium shots of big bearded men in conjunction with situations where you are otherwise unable to produce a beard or to make a fuller more realistic beard than SDXL usually produces.
Start at LoRa strength 0.4, work your way up to 0.7
https://ko-fi.com/alightbulb if you're feeling generous
Description
FAQ
Comments (5)
Really loved your LoRA! These kind of beard is super hard to prompt with SD.
Would you consider training a SD 1.5 version?
TY. It's bad in SDXL as well. I am hard core focusing on SDXL right now (questionable choice), and I don't have plans to make 1.5 versions of my models any time soon but I wouldn't say never.
diogodJan 19, 2024
@CruzFlesh I still try to do both versions. Is not really about the training, but the choosing the best epoch is what takes the most of my time. More than captioning I guess.