onnxruntime
Make Flash Attention work on Windows
#21015
Merged

Commits
  • start work
    aciddelgado committed 1 year ago
  • flash updated
    aciddelgado committed 1 year ago
  • update to 3.5 cutlass
    aciddelgado committed 1 year ago
  • Add /Zc:__cplusplus
    tianleiwu committed 1 year ago
  • update cutlass
    tianleiwu committed 1 year ago
  • Add code to use batch hook
    tianleiwu committed 1 year ago
  • update cgmanifest
    tianleiwu committed 1 year ago
  • limit max head size = 1024
    tianleiwu committed 1 year ago
  • Merge branch 'tlwu/fix_cutlass_msvc_build_error' into aciddelgado/flash_windows
    aciddelgado committed 1 year ago
  • make flash work
    aciddelgado committed 1 year ago
  • flash attn working and memeff too
    aciddelgado committed 1 year ago
  • minor
    aciddelgado committed 1 year ago
  • merge conflict
    aciddelgado committed 1 year ago
  • fix test failure
    aciddelgado committed 1 year ago
  • lint and clang
    aciddelgado committed 1 year ago
  • dont build 11
    aciddelgado committed 1 year ago
  • comments and benchmark script
    aciddelgado committed 1 year ago
  • lint
    aciddelgado committed 1 year ago
  • try transformers test
    aciddelgado committed 1 year ago
  • alibi flag
    aciddelgado committed 1 year ago
  • Merge branch 'main' into aciddelgado/flash_windows
    aciddelgado committed 1 year ago
  • fixes
    aciddelgado committed 1 year ago
  • lint exlcude
    aciddelgado committed 1 year ago
  • clang lint stuff
    aciddelgado committed 1 year ago
  • cpp lint sucks
    aciddelgado committed 1 year ago
  • no cpp lint this folder
    aciddelgado committed 1 year ago
Loading