New video demo: End-to-end MLOps using Pachyderm, Determined, and KServe

In this talk from ML-at-Scale 2023, Andrew Mendez shows how to use the PDK stack for data-driven, end-to-end model training pipelines. The PDK stack consists of these free and open-source libraries:

  • Pachyderm for data versioning and data-driven pipelines.
  • Determined for distributed training and hyperparameter search.
  • KServe for model serving.

Check out the video below!

Join the Pachyderm & Determined communities on Slack: