Jakevin's picture
Add 8bit/ (uniform 8-bit affine, group size 64; built with runner/run_mlx.py q8_g64) and update model card with measured results
ddabdbe verified