Skip to content

[PyTorch] MXFP4 weight QAT on MXFP8 and FP8 block-scaling recipes - #3264

Draft
xiuhu17 wants to merge 9 commits into
NVIDIA:mainfrom
xiuhu17:v4_mxfp4_qat
Draft

[PyTorch] MXFP4 weight QAT on MXFP8 and FP8 block-scaling recipes#3264
xiuhu17 wants to merge 9 commits into
NVIDIA:mainfrom
xiuhu17:v4_mxfp4_qat

Commits

Commits on Jul 25, 2026

Commits on Jul 26, 2026

Commits on Jul 27, 2026