These models are merged from style LoRAs into the himawarimix model.
Parameters used when training the LoRA models:
v2.1b:
-