Skip to content

[float8nocompile] Add alternate Triton kernels for FP8 conversion which use atomic_max-based algo instead of reduction-based algo #1074

[float8nocompile] Add alternate Triton kernels for FP8 conversion which use atomic_max-based algo instead of reduction-based algo

[float8nocompile] Add alternate Triton kernels for FP8 conversion which use atomic_max-based algo instead of reduction-based algo #1074

Annotations

1 warning

Check PR Labels

succeeded Dec 20, 2024 in 2s