Only quantization methods that can be run with PyTorch modules are currently supported.