I haven't done much training yet but you might try a textual inversion embedding instead of lora.
I've seen those but haven't read up on them. The Stable Diffusion world is pretty overwhelming!
Please Login to reply.
No replies yet.
