This LoRA is trained on a combined dataset of artists: Bartolomeobari (12 images), Wlop (12 images), AmaliMirror LoRA generations (12 images), as well as a small dataset (36 images) of randomly selected artists from Pinterest, each characterized by soft pastel colors and sharp character lines.
This LoRA is not one of those "secret LoRAs", but I use it as a base for soft color work without the use of additional tags and style settings.