CNN quantization using xilinx brevitas

Question

I have been trying to apply quantization to a CNN using brevitas Xilinx. However I cant get it working. I have not written the python code that creates the model so ideally I would like to apply post training quantization (PTQ), because I want to keep the provided code unchanged. So I want to ask if is there a way to get an onnx/keras/torch model and apply PTQ using brevitas?

I have been given the following parameters to work with:

--target_backend fx --scale_factor_type po2_scale --export_onnx_qcdq

but it is not clear to me how to use them. Any insights would be appreciated.