Trained on 25 images of Margot Robbie for 1600 steps :D.

Update: This model doesn’t seem to require a trigger word. I used "margotrobb1e", but the lora works regardless of its inclusion. I’m still exploring how SimpleTuner handles trigger words.

Below is the accompanying guide that shows my optimal settings and how to train on runpod.
https://civarchive.com/articles/6678