This lora is yet far from being perfect. To see which checkpoints and settings I used, you can load the pictures in ComfyUI.
# Output
Better chin shape
More detailed face
# Training
Changes to dataset
chzzksonycast__s0nyc4st__0_3.safetensors