Prerequisites

1. To enable net saving via HDF, copy `utils/dump_network_hdf5.cpp` to
   `<CAFFE_DIR>/tools` and compile
2. To enable net saving via tensor-flow, clone:
   `[email protected]:ethereon/caffe-tensorflow.git`

Test

After setting paths and options in each script:

$ ./convert.sh
$ python check_models.py

Model Conversion

Supports models based on VGG-16:

$ python convert_model.py