SnapML

Description

Snap ML is a library for training generalized linear models. It is being developed at IBM® with the vision to remove training time as a bottleneck for machine learning applications. Snap ML supports many classical machine learning models and scales gracefully to data sets with billions of examples or features. It also offers distributed training, GPU acceleration, and supports sparse data structures.

pai4sk

The pai4sk conda package includes the IBM® accelerated Machine Learning library. The main component of this library includes SnapML APIs. Snap ML is a library for training generalized linear models. It is being developed at IBM with the vision to remove training time as a bottleneck for machine learning applications. 

For how to use this framework visit the WML CE Knowledge Center Link

SnapML Spark

SnapML Spark APIs are shipped under the conda package snapml-spark. This package offers distributed training of models across a cluster of machines. SnapML APIs can be used in a SparkML machine learning pipelines.

For how to use this framework visit the WML CE Knowledge Center Link

Tags

The images are tagged in two ways. By framework version, and by WML CE version.  The framework versioning references the specific version of the framework included in the image.  While the WML CE tagging will tag the specific version of the framework that's included in a release of WML CE.  

Tag Format 1:

Tags images based on the specific project/frameworks version.

<framework-version>-<non-cpu|cpu>-<python-version>-<architecture>-<build-version>

• framework-version -- The version of the specific framework installed in the image
• non-cpu/cpu -- If cpu is specific in the tag, then the framework is designed to use cpu-only for execution.  The lack of the cpu tag implies the image is built for gpu based execution. (Not all frameworks have cpu versions)
• python-version -- Which version of python 3 that the framework was built against.
• architecture -- The computer architecture that the image is built with.  ppc64le for POWER architecture based images, and x86_64 for x86 based images.(Not all frameworks have x86_64 versions)
• build-version -- Version of the image build.  Increments starting at 1 for each rebuild of the particular image. 

Tag Format 2:

Tags images based on the version of the framework shipped with a specific wml-ce version.

wmlce-<wml-ce-version>-<non-cpu|cpu>-<python-version>-<architecture>

• wml-ce-version -- The version of wml-ce that the framework was pulled from.
• non-cpu/cpu -- If cpu is specific in the tag, then the framework is designed to use cpu-only for execution.  The lack of the cpu tag implies the image is built for gpu based execution.
• python-version -- Which version of python 3 that the framework was built against.
• architecture -- The computer architecture that the image is built with.  ppc64le for POWER architecture based images, and x86_64 for x86 based images.