Skip to content

GitLab

Switch branch/tag
Find file BlameHistoryPermalink
  • Jeff Daily's avatar
    enable ROCm build; add BF16 for ROCm and CUDA (#325) · 9a651d91
    Jeff Daily authored 
    * first step, everything compiles

* fix rebuilds; skip cuda version check for rocm

* use macro for __shfl_up_sync __shfl_down_sync

* add BFloat16 support for ROCm and CUDA

* add USE_ROCM definition to setup.py

* flake8 fixes
    9a651d91
setup.py 4.58 KB