Even though I already trained a LoRa that sorta do stop-motion style I decided to train another model from scratch since I wasn't satisfied with the results. This should work better and have a more distinct style
Use stopmotion as first token
tilt-shift adds a nice shallow depth of field effect that reinforces the style
put painting and drawing in the negative prompt to obtain cleaner results
use weight 1, lower it if it's too crowded
The best results requires a low CFG (7) and high steps number (I used 60), I know, this sucks
Hope you like it, let me know
if you like my LoRas just publish some image made with them to help me gain buzz
Training LoRa necessitate a lot of coffee. If you think that what I do is useful and want to help me keep doing this you can buy me a coffe here: https://ko-fi.com/clumsy_trainer