Steps to run this script

1. Transfer model description(Caffee, TensorFlow...etc) to representation of NNVM(.json).
2. Set hardware configure(hw_config.pl).
3. Input the representation and execute with Performance estimation tool(Estimater.pl).

• You can run:

   perl /path/to/Estimater.pl -f /path/to/model.json -o /path/to/Result.csv