AIF-C01 The primary purpose of a model card? Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  To provide a detailed performance benchmark on a single metric

  Why it's wrong here

  A model card documents intended use, limitations, training data, ethical considerations and evaluation results across multiple dimensions, not a single metric. It is tempting because performance figures appear on cards, but a card reporting only one benchmark would omit the transparency and risk information that is its actual purpose.

• ✗

  To register the model in SageMaker Model Registry

  Why it's wrong here

  Model cards are documentation artefacts describing a model's purpose, data, evaluation and limitations; they are not a registration mechanism. Registering a model in SageMaker Model Registry is a separate governance step, which tempts because both concern model lifecycle, but registration uses the registry API, not a card.

• ✓

  To document the model's intended use, performance, and limitations for transparency

  Why this is correct

  A model card records intended use, performance metrics, training data characteristics and known limitations, giving consumers the transparency needed to judge whether a model suits their context. It is documentation, not a runtime artefact, so it does not enforce guardrails or govern inference.

• ✗

  To store the model's parameters and weights for deployment

  Why it's wrong here

  Model cards hold descriptive documentation, not weights or parameters; those are stored in model artefacts such as Amazon S3 objects referenced by the model package. Storing weights is tempting because deployment requires them, but a card's role is to describe the model, not to carry its binary payload.

Go deeper

Related to this question