>>938806896
I have another workflow where it uses an abliterated Llama for prompt enhancement after auto tagging. I don't really know if that's better