Skip to content

Optimize quantized MoE and block-scaled FP8 CUDA projections - #33473

Draft
Akshay Sonawane (apsonawane) wants to merge 15 commits into
mainfrom
asonawane/main-ops-quantized-moe-fp8
Draft

Optimize quantized MoE and block-scaled FP8 CUDA projections#33473
Akshay Sonawane (apsonawane) wants to merge 15 commits into
mainfrom
asonawane/main-ops-quantized-moe-fp8

Fix MoE routing NaN race and CUDA plugin test linkage

fac1638
Select commit
Loading
Failed to load commit list.

Select a check to view from the sidebar