Repository navigation

[PyTorch] Propagate skip_fp8_weight_update in GroupedLinear during FP8 CUDA graph capture - #3065

Merged

ksivaman merged 3 commits into

NVIDIA:mainfrom

LeSingh1:fix-grouped-linear-skip-fp8-weight-update-3051

Jun 10, 2026

Merged

[PyTorch] Propagate skip_fp8_weight_update in GroupedLinear during FP8 CUDA graph capture#3065
ksivaman merged 3 commits into
NVIDIA:mainfrom
LeSingh1:fix-grouped-linear-skip-fp8-weight-update-3051

Commits on Jun 1, 2026
- [PyTorch] Propagate skip_fp8_weight_update in GroupedLinear during FP8 CUDA graph capture
  LeSingh1
  committed
  84575e8
- [PyTorch] Add CUDA graph FP8 weight-caching test for GroupedLinear
  LeSingh1
  committed
  6c75a8e
Commits on Jun 9, 2026
- Merge branch 'main' into fix-grouped-linear-skip-fp8-weight-update-3051
  ksivaman
  authored
  ed37453