Deep Learning (DL) · Advanced level

Advanced level · Module 6: Deep Learning Capstone · Lesson 4 of 5

Evaluation and Documentation

Evaluate the deployed model on field-like data and document it fully.

In this lesson you will

  • Evaluate across conditions and groups
  • Write the model card and risk assessment

Key concepts

field-like evaluationdisaggregated resultsmodel cardrisk assessment

The lesson

Evaluate the deployed model on field-like test data, broken down by conditions and groups.

Write the model card with intended use, performance, limits and prohibited uses, and update the risk assessment with red-team findings.

Document how to retrain, deploy and monitor the model.

Step by step

  1. Run the field-like evaluation.
  2. Run a short red-team exercise.
  3. Write the model card and risk assessment.
  4. Write the retraining and deployment guide.

Practice activity

Complete milestone 4: evaluation and documentation.

Expected result: An honest evaluation and complete documentation.

Check your understanding

  1. Why break down evaluation by conditions?

    Show answer

    To find where the model fails in the field.

  2. What should the documentation enable another team to do?

    Show answer

    Retrain, deploy and monitor the model.

Taught live by an instructor — online or in person. Free course lessons are for self-study and do not lead to a certificate — see registration options.