Experimental LoRA Trying to make a more flexible model, still probably overbaked.
Kyle.safetensors
440113_training_data.zip