Update flashinfer CUTLASS NVFP4 MoE Kernel to use per expert global scaling factor#21408

vllm-bot merged 2 commits into

vllm-project:mainvllm-project/vllm:mainfrom

wenscarl:flashinfer_moe_updatewenscarl/vllm:flashinfer_moe_updateCopy head branch name to clipboard