Track executions and artifacts

Vertex AI Experiments supports tracking both executions and artifacts. Executions are steps in an ML workflow that include but aren't limited to data preprocessing, training, and model evaluation. Executions can consume artifacts such as datasets and produce artifacts such as models.

Create artifact

The following sample uses the create method of the Artifact Class.

Python

def create_artifact_sample(
    schema_title: str,
    project: str,
    location: str,
    uri: Optional[str] = None,
    resource_id: Optional[str] = None,
    display_name: Optional[str] = None,
    schema_version: Optional[str] = None,
    description: Optional[str] = None,
    metadata: Optional[Dict] = None,
):
    artifact = aiplatform.Artifact.create(
        schema_title=schema_title,
        uri=uri,
        resource_id=resource_id,
        display_name=display_name,
        schema_version=schema_version,
        description=description,
        metadata=metadata,
        project=project,
        location=location,
    )
    return artifact
  • schema_title: Required. Identifies the schema title used by the resource.
  • project: Your project ID. You can find these IDs in the Google Cloud console welcome page.
  • location: See List of available locations.
  • uri: Optional. URI of artifact's location.
  • resource_id: Optional. The resource_id portion of the Artifact name with the format. This is globally unique in a metadataStore:
    projects/123/locations/us-central1/metadataStores/<metadata_store_id>/artifacts/<resource_id>.
  • display_name: Optional. The user-defined name of the resource.
  • schema_version: Optional. Specifies the version used by the resource. If not set, defaults to use the latest version.
  • description: Optional. Describes the purpose of the resource to be created.
  • metadata: Optional. Contains the metadata information that will be stored in the resource.

Start execution

The following sample uses the start_execution method.

Python

def start_execution_sample(
    schema_title: str,
    display_name: str,
    input_artifacts: List[aiplatform.Artifact],
    output_artifacts: List[aiplatform.Artifact],
    project: str,
    location: str,
    resource_id: Optional[str] = None,
    metadata: Optional[Dict[str, Any]] = None,
    schema_version: Optional[str] = None,
    resume: bool = False,
):
    aiplatform.init(project=project, location=location)

    with aiplatform.start_execution(
        schema_title=schema_title,
        display_name=display_name,
        resource_id=resource_id,
        metadata=metadata,
        schema_version=schema_version,
        resume=resume,
    ) as execution:
        execution.assign_input_artifacts(input_artifacts)
        execution.assign_output_artifacts(output_artifacts)
        return execution
  • schema_title: Identifies the schema title used by the resource.
  • display_name: The user-defined name of the resource.
  • input_artifacts: Artifacts to assign as input.
  • output_artifacts: Artifacts as outputs to this Execution.
  • project: Your project ID. You can find these in the Google Cloud console welcome page.
  • location: See List of available locations.
  • resource_id: Optional. The resource_id portion of the Artifact name with the format. This is globally unique in a metadataStore: projects/123/locations/us-central1/metadataStores/<metadata_store_id>/artifacts/<resource_id>.
  • schema_version: Optional. Specifies the version used by the resource. If not set, defaults to use the latest version.
  • metadata: Optional. Contains the metadata information that will be stored in the resource.
  • resume: bool.

    Note: When the optional resume parameter is specified as TRUE, the previously started run resumes. When not specified, resume defaults to FALSE and a new run is created.

Notebook samples