Skip to content

[ROCm][Quantization][Kernel] Use FP8 FNUZ when OCP flag is 0 or undefined #13851

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Feb 27, 2025
Merged

[ROCm][Quantization][Kernel] Use FP8 FNUZ when OCP flag is 0 or undefined #13851

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions csrc/quantization/fp8/amd/quant_utils.cuh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,12 +24,12 @@ __inline__ __device__ Tout scaled_vec_conversion(const Tin& x,
return x;
}

#if HIP_FP8_TYPE_FNUZ
using fp8_type = __hip_fp8_e4m3_fnuz;
using fp8x2_type = __hip_fp8x2_e4m3_fnuz;
#elif HIP_FP8_TYPE_OCP
#if HIP_FP8_TYPE_OCP
using fp8_type = __hip_fp8_e4m3;
using fp8x2_type = __hip_fp8x2_e4m3;
#else
using fp8_type = __hip_fp8_e4m3_fnuz;
using fp8x2_type = __hip_fp8x2_e4m3_fnuz;
#endif

// fp8 -> half
Expand Down