Presentation
BYOD Lab: Manage Life Cycle of Generative AI Model Deployment on zCX for OpenShift with IBM Spyre
Contributors
Event Type
Technical Session
Machine Learning/AI
z/OS Systems Programming
All Audiences
Integrate with Distributed Systems and Cloud
TimeThursday, August 202:30pm - 3:30pm EDT
LocationRoom 316 BYOD LAB
DescriptionThis lab would demonstrate the end-to-end AI application lifecycle on a zCX for Red Hat OpenShift cluster using Red Hat OpenShift AI and Spyre accelerator(s). It walks through creating a data science project using Jupyter Notebook, deploying the Granite 3.3 8B Instruct model, serving the model using Kserve, and utilizing the vLLM runtime, with IBM Spyre accelerator(s). In addition, the session will further demonstrate running inference queries and analyzing model logs to validate system behavior and performance. This integrated approach enables data scientists, developers, and operations teams to collaborate on a co-located single platform, eliminating silos and reducing handoff delays across the AI lifecycle.
Contributors
zCX Development Lead
IBM STSM
Senior AI Client Engagement Leader