Use firedotinc for the activation tag.
This is a lora trained to reproduce the style of... My own work. Trained on noob vpred 1.0
This is also effectively a dataset test/proof of concept, hence why it's a miniscule 1dim LoRA instead of my regular, fat GLoRA trains. Which is to say, it's only 1dim - it's only 5.36MB! It's completely functional but I wouldn't say it's as faithful in reproduction as it could be due to being completely starved of DIM. But it's still pretty okay for what it is.
I should have an actual EPS GLoRA uploaded in the next few days or so - however I don't have plans on doing one for VPRED so this is all it'll get.
You can find a 2dim lora trained on noob 0.5 EPS here