This is the same LORA but trained over Anything V3 so seems to work better with mixes and trained models
This is the same LORA but trained over Anything V3 so seems to work better with mixes and trained models