matrixglitch

/

SD3_prompt-llama_8b

like 3

I looked @ your datasets, I presume it is Magic Prompt L but I wanted to confer with ya due to the #'s in the readme differing from what I see in it. I've never done a LoRA or FT of a model, but I know I'll end up wanting to do this w/ a AQLM 2-bit Llama or maybe a 4-bit 70B Llama (=

Cheers and thank you for sharing all these gems w/ us!