pub fn kquant_mmvq_experts_w(
gpu: &dyn GpuBackend,
kernel: KernelHandle,
vxs: DevicePtr,
y_q8: DevicePtr,
out_bf16: DevicePtr,
n: u32,
k: u32,
m: u32,
n_experts: u32,
y_stride_bytes: u32,
stream: u64,
) -> Result<()>Expand description
kquant_mmvq_*_experts_w: the warp-per-row expert batch, same arguments
as kquant_mmvq_experts.