This guide demonstrates how to use TensorFlow Graph Transform Tool to optimize a frozen TF graph before deploying it in production. There are different types of optimizations. One is to make the model smaller and faster in size without accuracy loss. And the other is to change the weights from higher precision to lower precisions, usually from FP32 to FP16 or INT8. For deploying your model to a phone or embedded device, you can optimize away batch normalization or other training-only features. The TensorFlow Graph Transform framework offers a suite of tools for modifying computational graphs, and a framework to make it easy to write your own modifications.