2 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
tql.99and戚余航	3f2e315f6e	optimize: reduce shulffle and quantization overhead in cutlass_moe sm90 (#8962) ... Co-authored-by: 戚余航 <qiyuhang@bytedance.com>	2025-08-09 00:29:12 -07:00
tql.99	e67276ecb3	feat: support cutlass_moe_fp8 kernel for fusedmoe in sm90 (#8678)	2025-08-03 10:47:15 -07:00