Mojo function

multistage_qgemm_kernel

multistage_qgemm_kernel[c_type: DType, c_layout: Layout, a_type: DType, a_layout: Layout, b_packed_type: DType, b_layout: Layout, group_size: Int, pack_factor: Int, transpose_b: Bool, config: MatmulConfig[a_type, b_packed_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})](c: LayoutTensor[c_type, c_layout, MutableAnyOrigin], a: LayoutTensor[a_type, a_layout, MutableAnyOrigin], b_packed: LayoutTensor[b_packed_type, b_layout, MutableAnyOrigin])