Keras+theano+tensorflow+darknet

1. Keras Installation:
   It is best to build in the Anaconda virtual environment:
   Conda create-n Environment Name python=3.6
   Enter the environment:
   Source Activate Environment name
   Install Keras:
   Pip Install Keras
   Pip Install Theano
   Pip Install tensorflow-gpu==1.2.0
   If you use Theano as backend, you need to Conda install PYGPU to support parallel and gou operations.

If Modulenotfounderror:no module named ' Mkl ' appears
To demote the MKL in the current environment to 2017:
Conda Uninstall MKL
Conda Install mkl==2017.0.4

1. Caffe Model Transfer Keras: Subsequent updates

Keras+theano+tensorflow+darknet