Trained using the Rapid-Engine and using the same dataset as my SD1.5 LoRA. Tends to exaggerate her age a bit which can be mitigated to a degree by using specific prompts and/or modifying LoRAs.
First release