This repo is a set of GGUF quants of a grimjim/Magnolia-v3b-12B, a merge of pre-trained language models created using mergekit.
llama.cpp was used to make the following quants:
4-bit
5-bit
6-bit
8-bit
Base model
