OpenShift Commons Briefing: Introduction to MLFlow on OpenShift - Mani Parkhe (Databricks) and Zak Hassan (Red Hat)

In this OpenShift Commons Briefing, DataBricks' Mani Parkhe gave an excellant introduction to MLFlow, an open source platform to manage the Machine Learning lifecycle, including experimentation, reproducibility and deployment.

MLFlow currently has three components:

• MLflow Tracking - Record and query experiments: code, data, config, and results.
• MLflow Projects -Packaging format for reproducible runs on any platform.
• 
MLflow Models - General format for sending models to diverse deployment tools.

Following the introduction, Red Hat's Zak Hassan gave an demonstration of the MLFlow operator for deploying MLFlow on Kubernetes.  Building machine learning systems is not just a one-off effort; the process is iterative, and therefore, managing the lifecycle of the machine learning algorithms and applications is a key factor for success. Naturally, an operator for MLFlow to helps  it better integrate into a Kubernetes environment.

Databricks MLFlow Intro Slides from the briefing: MLflow at OpenShift Commons ML SIG meeting Feb 2019

Red Hat's Zak Hassan's MLFLow Operator Slides from the briefing: OpenShift Commons ML Sig - Experiment Tracking On OpenShift With MLFLow

Additional Resources:

• https://mlflow.org
• https://github.com/mlflow
• https://github.com/mlflow/mlflow/pull/932
• https://github.com/zmhassan/mlflow-tracking-operator
• https://www.operatorhub.io/what-is-an-operator
• https://www.operatorhub.io/contribute

Join us for the Next OpenShift Commons Machine Learning SIG meeting

Check out the upcoming SIG meeting dates on the calendar here: https://commons.openshift.org/events.html

Join the Community at the Upcoming OpenShift Commons Gathering in Silicon Valley! March 11th @ The Hyatt Regency in Santa Clara, California

The OpenShift Commons Gathering brings together experts from all over the world to discuss the container technologies, operators, the operator framework, best practices for cloud-native application developers and the open source software projects that underpin the OpenShift ecosystem to help take us all to the next level in cloud-native computing. This next gathering will feature 400+ developers, project leads, cloud architects, DevOps professionals, sysadmins, and cloud-native practitioners coming together to explore the next steps in making container technologies successful and secure at scale.

Confirmed Keynotes and Speakers from Red Hat already include:

• AMA Panel with OpenShift Product Managers and Engineering leads
• Rob Szumski on Kubernetes Operators
• Reza Shafii on Red Hat's Unified Hybrid Cloud
• Brian “redbeard” Harrington on the State of Serverless on Kubernetes
• Diane Mueller on Cross-Community Collaboration with Upstream
• Future of Linux Cgroups, Container Security, and more..
• Speakers from Google, Facebook, Splunk, Rook.io and Uber

More speakers and panelists are being added, check out the agenda for updates. Please note: Pre-registration is required. Click here to register.

Join OpenShift Commons

If you'd like to connect and participate in the OpenShift Commons, the open source community for members of the OpenShift ecosystem - sign up and join here!