Pytorch Mobile allows for integration of QNNPACK for 8-bit quantized kernels.