Description: An inference can be directly embedded within a C++ or C program through the C API and linking against TensorFlow libraries. Alternatively, the inference may be performed from a model server running on the same compute node as the target program. This course will consist of a two-part tutorial covering these multiple approaches to embed TensorFlow models in C++ programs. This will include demonstration of model saving and preparation within Python, an overview of compilation of necessary libraries on the HPCs, launching model servers, and developing model inference functions within C++. Advanced topics like coupling MPI with multi-GPU inferencing will also be discussed.

Presenter(s): Dr. Mathew Boyer, GDIT / PET
Location: Webcast
Date & Time: February 15 & 17, 2022, 2:00 - 3:30p ET

Distribution Statement D. Distribution authorized to the Department of Defense and U.S. DoD contractors only, Administrative or Operational Use, 15 February 2022.  Other requests for this document shall be referred to the High Performance Computing Modernization Office, 3909 Halls Ferry Road, Vicksburg, MS 39180.