Business Strategies for Generative AI Solutions →hardMultiple Choice

Generative AI Leader Practice Question: Business Strategies for Generative AI Solutions

Exhibit

Refer to the exhibit.

```
{
  "displayName": "my-pipeline",
  "pipelineSpec": {
    "root": {
      "inputDefinitions": {},
      "task": {
        "componentRef": {
          "name": "comp-model-eval"
        },
        "inputs": {
          "project": {
            "runtimeValue": {
              "constantValue": "my-project"
            }
          },
          "location": {
            "runtimeValue": {
              "constantValue": "us-central1"
            }
          },
          "model_name": {
            "componentInput": "model_name"
          },
          "eval_dataset": {
            "componentInput": "eval_dataset"
          }
        }
      }
    }
  },
  "runtimeConfig": {
    "parameterValues": {
      "model_name": "text-bison@001",
      "eval_dataset": "projects/my-project/datasets/eval"
    }
  }
}
```

A machine learning engineer is defining a Vertex AI pipeline for model evaluation using the JSON representation shown. The pipeline fails with an error that the 'eval_dataset' parameter is missing. What is the issue?

⚠ Common exam trap

In Google Vertex AI pipelines, a common trap is confusing the component's input definition with the pipeline's input parameter declaration. The pipeline must explicitly declare all inputs in `pipelineSpec.root.inputDefinitions.parameters`, not just in the component spec.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

The pipeline spec does not declare 'eval_dataset' as a pipeline input parameter

The pipeline fails because the JSON representation of the Vertex AI pipeline does not include 'eval_dataset' in the `pipelineSpec.root.inputDefinitions.parameters` section. Without declaring it as a pipeline input parameter, the pipeline runtime cannot resolve the reference to `inputs.eval_dataset` in the component's arguments, causing the missing parameter error.

Answer analysis

Option-by-option breakdown

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

✗
The component 'comp-model-eval' does not accept 'eval_dataset' as input
Why it's wrong here
The component's declared inputs define what the pipeline can pass; if 'eval_dataset' is absent from that interface, the runtime argument is rejected as unknown. Components are correct when their input schema already lists the parameter, so the pipeline merely supplies a value. Here the interface itself omits it.
✗
The 'project' parameter should be a pipeline input, not a constant
Why it's wrong here
Whether 'project' is a pipeline input or a hard-coded constant does not affect 'eval_dataset' resolution. Constants are legitimate for stable values such as project IDs. Making 'project' an input would be correct when the same pipeline must run across multiple projects, not for this missing-parameter error.
✗
The runtimeConfig parameter values must be strings, not references
Why it's wrong here
runtimeConfig values accept strings, integers, floats, booleans and references; parameter references are valid there. The failure concerns a missing component input, not value typing. String-only runtimeConfig would be the issue when a non-string literal is passed directly, which is not this scenario.
✓
The pipeline spec does not declare 'eval_dataset' as a pipeline input parameter
Why this is correct
Vertex AI pipelines resolve parameters declared in the pipeline spec's inputs section; a value passed at runtime cannot bind to an undeclared name. Because 'eval_dataset' never appears as a pipeline input parameter, the component's reference fails resolution, producing the missing-parameter error.

Visual reference

Go deeper

Generative AI Leader Practice Question: Business Strategies for Generative AI Solutions

Exhibit

Why each option matters

Option-by-option breakdown

Related to this question