liveBook · Manning

Step annotation

As explained in Designing Deep Learning Systems: A guide for software engineers

A Metaflow annotation that marks a function as a workflow step.

A step annotation marks a function as a Metaflow workflow step. Annotated functions, connected with self.next, construct a workflow DAG directly in application code, keeping prototyping and workflow development together. section 9.3.3

Use annotated steps to:

  • Represent sequential workflow stages without defining a separate DAG system. section 9.3.3
  • Test the same workflow locally before exporting it to AWS Step Functions or Argo Workflows. section 9.3.3

How Designing Deep Learning Systems: A guide for software engineers covers it

How deep it goes
Defines annotation and step connections with a runnable workflow pattern, but does not explain annotation internals.
Examples
Uses data fetching, transformation, database storage, team notification, fan-out, and local-to-production deployment commands.
What it assumes
Assumes Python class-based workflow code and Metaflow’s FlowSpec, while demonstrating step connections from a working example.
What you can do after
Build it from scratch by annotating functions, connecting them with self.next, testing locally, and exporting the resulting workflow.
How it connects
Connects workflow DAGs, workflow orchestration, local testing, AWS Step Functions, and Argo Workflows.
Cover of Designing Deep Learning Systems: A guide for software engineers
liveBookmanning.com
Read Designing Deep Learning Systems: A guide for software engineers on liveBook Get the book at manning.com

This explainer was generated by Manning's knowledge graph from the text of Designing Deep Learning Systems: A guide for software engineers. The section links point to the parts of the book it draws on.