XNNPACKQuantizer.set_module_name() not working as expected

quantization
4

Thank you. So does that mean it’s properly quantized? I still wonder why quantizer.set_global(qconfig) reduces size of the weights.