This is a llama13b qlora merge I trained on the supercot dataset curated by kaiokendev.

I'll add more to the model card later.