JohnQinAMDandClaude Opus 5 ef5f7cd119 [ROCm][DSv4.1][Perf] Dequantize the MXFP8 weight once when dot_scaled cannot be used (#56560) ...

Signed-off-by: JohnQinAMD <yanyuan.qin@amd.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>

2026-09-15 09:07:26 +08:00

..

attention

[5/N][warmup][DSv4] Migrate NVIDIA CuTeDSL attention kernels (#53566)

2026-09-12 00:05:59 -04:00

linear

[ROCm][DSv4.1][Perf] Dequantize the MXFP8 weight once when dot_scaled cannot be used (#56560)

2026-09-15 09:07:26 +08:00

mhc

[ROCm][DSV4.1][Perf] Fold the mHC post step into the delayed pre projection (#56513)

2026-09-14 16:44:56 -05:00

__init__.py

[Refactor] [1/N] Reorganize kernel abstraction directory (#34055)

2026-02-24 06:47:22 +00:00