/flyte-pipeline

Superwise's webinar for setting up continuous training infrastructure

Primary LanguagePythonMIT LicenseMIT

Superwise Webinar - Continous ML Stack

🚧 THIS IS FOR DEMONSTRATION PURPOSE ONLY 🚧

🚧 DO NOT USE IT IN PRODUCTION 🚧

3 repos:

  • mlflow_server - for deploying mlflow using docker-compose (metadata store + model registry)

  • inference service - simple flask application for real-time inference

  • flyte_pipeline - setting up flyte cluster locally and running training-serving pipeline on top

How To Start

  • Overwrite the file envs.env with your values.
  • Export the vars into your current terminal session source envs.env.
  • Set up mlflow server by reading mlflow_server/README.md

FEEL FREE TO REACH OUT AND TALK TO US - (SUPERWISE.AI)

HAPPY MACHINE LEARNING !