auto-round
23ba0dbf - test: add int4-sym DPAS vs scalar token-count sweep for MoE decode threshold tuning

Commit

13 days ago

test: add int4-sym DPAS vs scalar token-count sweep for MoE decode threshold tuning Co-authored-by: a32543254 <53296245+a32543254@users.noreply.github.com>

References

#2111 - Optimize INT4 MoE prefill/decode (sym) XPU kernels with dedicated w4a16 tiling

#2143 - feat: W4A8 ARK XPU MoE kernel (int4 weight / int8 compute) with prefill + decode

Author

Copilot

Parents

87b617ea

auto-round 23ba0dbf - test: add int4-sym DPAS vs scalar token-count sweep for MoE decode threshold tuning

auto-round
23ba0dbf - test: add int4-sym DPAS vs scalar token-count sweep for MoE decode threshold tuning