Reproducing and scaling machine learning experiments with Polyaxon

Building modern machine learning systems involve various components and often requires connecting and managing different services, which introduces huge barriers of complexity in adopting machine learning. Infrastructure engineers and data scientists will often spend a significant amount of time manually tweaking deployments. https://polyaxon.com is an extensible open source machine learning stack on Kubernetes that addresses these concerns.

Transcript

1. Reproducing and scaling machine learning experiments with Polyaxon
2. Overview • Introduc9on • experimenta9on workflow • Polyaxon users • Demo • Road map and Q&A
3. Information • Name: Mourad Mourafiq • twiEer: @mmourafiq • GitHub: mouradmourafiq
4. What is polyaxon • Simplifies machine learning experimenta9on workflow • Can be deployed on premise or on cloud plaLorm • Solves issues related to reproducibility • Open source • Works with any library or framework • Can be used by single users or large organiza9ons
5. Experimentation workflow • Managing code, data, params, ar9facts • Scheduling and retraining experiments • Sharing and reproducing results • Deploying models !5
6. Polyaxon users • Direct • machine learning engineers • data scientists • data analytics engineers • Indirect • dev ops • developers
7. Demo
8. Roadmap and Q&A