pub fn launch_k3_kda_decode_token_on_device(
gpu: &dyn GpuBackend,
kernels: &K3KdaDecodeKernels,
x_qkv: &[f32],
conv_w: &[f32],
gate: &[f32],
beta: &[f32],
cfg: &KdaConfig,
device: &KdaDeviceState,
stream: u64,
) -> Result<Vec<f32>>Expand description
One-token CUDA KDA on device-resident conv/recurrent. Does not D2H state.