Mojo function

grouped_matmul_kernel_sm100

grouped_matmul_kernel_sm100[a_type: DType, b_type: DType, c_type: DType, a_layout: Layout, b_layout: Layout, a_tile_layout: Layout, b_tile_layout: Layout, c_layout: Layout, block_tile_shape: IndexList[3], mma_shape: IndexList[3], a_desc_layout: Layout, b_desc_layout: Layout, a_swizzle: TensorMapSwizzle = TensorMapSwizzle.SWIZZLE_128B, b_swizzle: TensorMapSwizzle = TensorMapSwizzle.SWIZZLE_128B, c_swizzle: TensorMapSwizzle = TensorMapSwizzle.SWIZZLE_NONE, transpose_b: Bool = True, num_threads: Int = 128, elementwise_lambda_fn: OptionalReg[elementwise_epilogue_type] = None](a_tma_op: TMATensorTile[a_type, a_tile_layout, a_desc_layout], b_tma_op: TMATensorTile[b_type, b_tile_layout, b_desc_layout], a_offsets: NDBuffer[DType.uint32, 1, MutAnyOrigin], expert_ids: NDBuffer[DType.int32, 1, MutAnyOrigin], c: LayoutTensor[c_type, c_layout, MutAnyOrigin], num_iters: Int)