Installing the MLDL frameworks

Find instructions for installing the machine learning and deep learning (MLDL) framework on an IBM® Power® System server.

Setting up the software repository

The PowerAI ML/DL packages are distributed as Conda packages in an online Conda repository. Conda must be configured to give priority to installing packages from this channel.

Add the PowerAI channel to the Conda configuration by running the following command:

conda config --prepend channels \
https://public.dhe.ibm.com/ibmdl/export/pub/software/server/ibm-ai/conda/

Creating Conda environments (optional)

With Conda, you can create environments that have different versions of Python and/or packages installed in them. Switching between environments is called activating the environment.

The syntax to create and activate a Conda environment is:

conda create --name <environment name> python=<python version>
conda activate <environment name>

The only valid Python versions with PowerAI are Python 2.7 and 3.6.

For example, to create an environment named powerai_env with Python 3.6:

conda create --name powerai_env python=3.6

conda activate powerai_env

For more information about what you can do with Conda environment, see https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html

Conda environments are optional; if not used, packages are installed in the default environment called 'base'

Installing all frameworks at the same time

All the ML/DL frameworks can be installed at the same time by using the powerai meta-package

With the Conda environment you want to install in activated run:

conda install powerai

Note: The Python 2.7 or 3.6 version of the package installed is determined by the Python version of the active environment.

Installing frameworks individually

You can install the ML/DL frameworks individually. The framework packages include the following versions:

Table 1. Framework packages
Package	Description	Version
`caffe`	IBM Optimized version of Berkeley Vision and Learning Center Caffe	1.0.0
`ddl`	Distributed Deep Learning	1.3.0
`pytorch`	PyTorch	1.0.1
`tensorflow-gpu`	TensorFlow with GPU support	1.13.1
`pai4sk`	PowerAI Snap ML	1.3.0
`snapml-spark`	PowerAI Snap ML Spark	1.2.0
`cudf`	Rapids cuDF	0.2.0
`cuml`	Rapids cuML	0.2.0

With the Conda environment activated, run the following command:

conda install <package name>

Accepting the PowerAI license agreement

During the Conda install, the packages will be downloaded from the internet and after downloading, the license agreement will be presented. Read the license agreement and accept the terms and conditions to complete the install. If you decline the license agreement the packages will not be installed.

After you finish reading the license agreement, future installs can be automated to silently accept the license agreement by running:

export IBM_POWERAI_LICENSE_ACCEPT=yes

before running the Conda install command.

The license accept has to be done only once on a per user basis.