There was no Bennett SDXL LoRa, so I made one.
Trained on 50 danbooru images in 5 epochs and 1250 steps, iirc, took only an hour.
This LoRa most likely won't work as intended with Pony Diffusion checkpoints, only SDXL.
Let me know if I should make one for Pony, although there are already some available.
TRIGGERWORD: bennett
accurate face generation:
goggles, goggles on head, white hair, short hair, hair between eyes, green eyes
Use tags like
white shirt, sleeveless, navel cutout, brown gloves, brown belt, belt pouch, blue shorts, goggles on head, boots
for original costume.
LICENSE
-none-
Do whatever you want, idc.
Except Stable Diffusion fair use, blah blah, whatever.
Have fun!