This is an experiment
Results will probably not be very good. I’m trying to figure out what types of images work well before I train a LoRA on my own cockatiel.
411335_training_data.zip
Cockatiel.safetensors