muralikrishnangm / tensorflow-c-example

Sample script for loading and inferencing a TensorFlow model in C

Introduction

Sample script for loading and inferencing TensorFlow models in C.

Author: Muralikrishnan Gopalakrishnan Meena (Oak Ridge National Laboratory), https://sites.google.com/view/muraligm/

Contributors:

• Murali Gopalakrishnan Meena

Description

These are sample scripts to load a TensorFlow model and run a forward step. See TensorFlow for C for TF C API installation instructions.

The following examples are provided (will be updated accordingly):

1. example_tensorflow_HelloWorld.c: A sample Hello World implementation
2. example_tensorflow_NNmodel.c: A simple model to replicate an activation function
   • Based on https://github.com/AmirulOm/tensorflow_capi_sample.git
3. example_tensorflow_tanh.c: A simple feedforwad NN for regression
   • Model based on https://github.com/muralikrishnangm/tutorial-ai4science-fluidflow#example-1
4. example_tensorflow_AICT.c: To load and perform inference on a CNN model. Used for the 2022 AAPM TrueCT Challenge

Usage

First save any TensorFlow or Keras model in your python script as:

my_model.save('model_dir_name', save_format='tf')

1. Activate binaries using environment file corresponding to your machine:

   source build/<machineName>_<cpu/gpu>.env
   

   Example: For running on CPU on Summit

   source build/summit_cpu.env
   

2. Compile: Update Makefile to select which model to do inference (change SRC in the file). Then do:

   make
   

3. Run:

   ./example.exe <options>
   

• Sample output for example_tensorflow_HelloWorld.c:

  Hello from TensorFlow C library version 2.4.1