This is the second round of testing for LORA using 20 training images!Include these prompts at a minimum :avian head, beak, egyptian, red eyes, dark skin, colored skin, helmet, bracelet, anklet,negative :card,For some reason, it tends to generate frames that look like social games, but this can be avoided by using a negative prompt.

Files with SHA256: 0174d8739ca5d83bd8dc619f5c82948490f3e5fea14007576f37c999abf99226