My firs LoRa, apologize if it's not perfect. All training suggestions are welcome.
Trained on 110 pictures, base models
recommendations for SD 1.5
EULER A
higher step number (25-50)
CFG 6-7
768x512
recommendations for SD 2.1 768x (don't forget to download .yaml)
EULER A
low step number (around 20-30)
CFG 6 or lower
960x576 or 1024x768
decrease LoRa weight if needed