I find it in mxnet/example/ssd, the symbol contains python files that are correspondent to json files of the pre-trained model, and I would like to know if it is achievable to convert the json file of the model to a python file.
Thank you for your reply.
From mxnet/example/ssd and mxnet/example/rcnn, there are symbol folder and the python code that describes certain network, and we can also find corresponding network parameters in json and params. However, from model_zoo, I can only find json and params files but the corresponding python code, at the moment, I use get_summary to print network structure, but it might be easier if this conversion can be done as well. Thanks,
You can certainly load the JSON model to create an in-memory Python object that you can then manipulate.
Thank you for the reply. I just gave a try, and it works as you suggested. Really appreciate your help!