another badly trained lora, Trained on NAI

placing dark-skinned male, eyepatch, facial hair, beanie, vest in the prompts may help

difference between v1 and light mode is that light mode has lower learning rate and network dim and alpha