[PyTorch] Propagate FP8 graph weight update flag in GroupedLinear#3052

ksivaman merged 3 commits into

NVIDIA:mainNVIDIA/TransformerEngine:mainfrom

allenphilipj:codex-grouped-linear-fp8-cudagraph-skipallenphilipj/TransformerEngine:codex-grouped-linear-fp8-cudagraph-skipCopy head branch name to clipboard