Mojo function

multistage_gemm

multistage_gemm[c_type: DType, c_shape: DimList, a_type: DType, a_shape: DimList, b_type: DType, b_shape: DimList, //, *, transpose_b: Bool, config: MatmulConfig[a_type, b_type, c_type, transpose_b], elementwise_lambda_fn: OptionalReg[fn[DType, Int, Int](IndexList[2], SIMD[$0, $1]) capturing -> None] = OptionalReg[fn[DType, Int, Int](IndexList[2], SIMD[$0, $1]) capturing -> None]({:i1 0, 1}), serial_reduction: Bool = False](c: NDBuffer[c_type, 2, origin, c_shape], a: NDBuffer[a_type, 2, origin, a_shape], b: NDBuffer[b_type, 2, origin, b_shape], runtime_config: MatmulConfig[a_type, b_type, c_type, transpose_b], ctx: DeviceContext)