Add blocksize=64 4-bit quantization support for ROCm CDNA (warp64) GPUs#1856

matthewdouglas merged 2 commits into

bitsandbytes-foundation:mainbitsandbytes-foundation/bitsandbytes:mainfrom

Abdennacer-Badaoui:rocm-64-blocksize-supportAbdennacer-Badaoui/bitsandbytes:rocm-64-blocksize-supportCopy head branch name to clipboard