[moe_training] Fused dual-axis MXFP8 quantization with correct metadata - #4995

[moe_training] Fused dual-axis MXFP8 quantization with correct metadata#4995

vkuzo merged 2 commits into

mainpytorch/ao:mainfrom

fused-mxfp8-dual-axis-quantizationpytorch/ao:fused-mxfp8-dual-axis-quantizationCopy head branch name to clipboard