Fix int8 zero-point overflow and conv bias_scale in eager quantized ref kernels#20655

meta-codesync[bot] merged 1 commit into

pytorch:mainpytorch/executorch:mainfrom

abeakkas:export-D110220645abeakkas/executorch:export-D110220645Copy head branch name to clipboard