kubecon24/content/day2/04_sponsored_ai_platform.md

655 B

title weight tags
Sponsored: Build an open source platform for ai/ml 4
keynote
ai

{{% button href="https://youtu.be/r7qd0ZHt4uE" style="warning" icon="video" %}}Watch talk on YouTube{{% /button %}}

Jorge Palma from Microsoft with a quick introduction.

Baseline

  • Kubernetes is cool and all
  • Challenges:
    • Containerized models
    • GPUs in the cluster (install, management)

Kubernetes AI Tool chain (KAITO)

  • Kubernetes operator that interacts with
    • Node provisioner
    • Deployment
  • Simple CRD that describes a model, infra and have fun
  • Creates inference endpoint
  • Models are currently 10 (Hugginface, LLMA, etc.)