Could you provide the original HF model (the .safetensors and etc) for more/better conversions? This would allow people to make other quants such as Q6_K_L, Q4_K_L, etc, add an imatrix (unless one already was, but my impression is no?) and so on.
· Sign up or log in to comment