GGUF repo for imatrix + some standard quants of Naphula/Goetia-26B-A4B-v1.7.
using imatrix generated on bartowski's combined dataset on q8_0 weights with -c 512 -b 512 -ub 512
quantized with llama.cpp #1692f9e50
nohurry SillyTavern presets strongly recommended for creative applications