-
Notifications
You must be signed in to change notification settings - Fork 40
Installing MaTEx TensorFlow CPU
Vinay Amatya edited this page Jun 22, 2017
·
16 revisions
We have provided a set of scripts for easy installation on CPU clusters. We expect that minimal changes would be required to the script for installation.
For using CPU enabled TensorFlow, you would need to set up the MPI_HOME and PYTHONHOME environment variables to point where MPI and Python 3.4 home directory should be found.
As an example, for bash shells:
$ export MPI_HOME=/where/mpi.h/lives
$ export PYTHONHOME=/where/tensorflow/packages/will/be/installed
Installation for bash shells
$ cd matex/src/deeplearning
$ cd cpu/py3.x
$ source ./install_mpi_tf.sh
Installation for C-shells
$ cd matex/src/deeplearning
$ cd cpu/py3.x
$ source ./install_mpi_tf.csh
This ensures that you are in a virtual python environment that encapsulates the distributed TensorFlow changes. Your shell prompt should look differently.
[py_distro] $
Getting Started on MaTEx-TensorFlow
- Required Software
- Installing MaTEx-TensorFlow on CPU Clusters
- Installing MaTEx-TensorFlow on GPU Clusters
- MaTEx-TensorFlow on Older glibc(v<2.19)
- DataSet Reader
- Testing Scripts
- Performance
- Running on PNNL Systems
- Running on NERSC Systems
- Restarting the MaTEx TensorFlow environment