Create and manage experiment runs

Use the Vertex AI SDK for Python to create and manage your experiment runs. You can use the Google Cloud console to delete experiment runs.

Vertex AI SDK for Python

The following samples use the methods init, start_run, and end_run from the aiplatform Package functions, and delete from the ExperimentClass.

Create and start run

Python

from typing import Optional, Union

from google.cloud import aiplatform


def create_experiment_run_sample(
    experiment_name: str,
    run_name: str,
    experiment_run_tensorboard: Optional[Union[str, aiplatform.Tensorboard]],
    project: str,
    location: str,
):
    aiplatform.init(experiment=experiment_name, project=project, location=location)

    aiplatform.start_run(run=run_name, tensorboard=experiment_run_tensorboard)
  • experiment_name: Provide the name of your experiment. You can find your list of experiments in the Google Cloud console by selecting "Experiments" in the section nav.
  • run_name: Specify a run name to associate with your current session. See start_run in the Vertex AI SDK reference documentation.
  • experiment_run_tensorboard: Optional. A backing TensorBoard resource to enable and store time series metrics logged to this experiment run using log_time_series_metrics.
  • project: Your project ID. You can find these IDs in the Google Cloud console welcome page.
  • location: See List of available locations

End run

Python

from google.cloud import aiplatform


def end_experiment_run_sample(
    experiment_name: str,
    run_name: str,
    project: str,
    location: str,
):
    aiplatform.init(experiment=experiment_name, project=project, location=location)

    aiplatform.start_run(run=run_name, resume=True)

    aiplatform.end_run()
  • experiment_name: Provide the name of your experiment. You can find your list of experiments in the Google Cloud console by selecting "Experiments" in the section nav.
  • run_name: Specify a run name.
  • project: Your project ID. You can find these in the Google Cloud console welcome page.
  • location: See List of available locations

Resume run

Python

from google.cloud import aiplatform


def resume_experiment_run_sample(
    experiment_name: str,
    run_name: str,
    project: str,
    location: str,
):
    aiplatform.init(experiment=experiment_name, project=project, location=location)

    aiplatform.start_run(run=run_name, resume=True)
  • experiment_name: Provide the name of your experiment. You can find your list of experiments in the Google Cloud console by selecting "Experiments" in the section nav.
  • run_name: Specify name of run that you want to resume.
  • project: Your project ID. You can find these in the Google Cloud console welcome page.
  • location: See List of available locations

Delete run

Python

from typing import Union

from google.cloud import aiplatform


def delete_experiment_run_sample(
    run_name: str,
    experiment: Union[str, aiplatform.Experiment],
    project: str,
    location: str,
    delete_backing_tensorboard_run: bool = False,
):
    experiment_run = aiplatform.ExperimentRun(
        run_name=run_name, experiment=experiment, project=project, location=location
    )

    experiment_run.delete(delete_backing_tensorboard_run=delete_backing_tensorboard_run)
  • experiment: The name or instance of this experiment. You can find your list of experiments in the Google Cloud console by selecting "Experiments" in the section nav.
  • run_name: Specify name of run that you want to delete.
  • project: Your project ID. You can find these in the Google Cloud console welcome page.
  • location: See List of available locations
  • delete_backing_tensorboard_run: Whether to delete the backing Vertex AI TensorBoard run that stores time series metrics for this run.

Manage status

Python

from typing import Union

from google.cloud import aiplatform


def update_experiment_run_state_sample(
    run_name: str,
    experiment: Union[str, aiplatform.Experiment],
    project: str,
    location: str,
    state: aiplatform.gapic.Execution.State,
) -> None:
    experiment_run = aiplatform.ExperimentRun(
        run_name=run_name,
        experiment=experiment,
        project=project,
        location=location,
    )

    experiment_run.update_state(state)
  • run_name: run name associated with your experiment
  • experiment_name: name of your experiment. You can find your list of experiments in the Google Cloud console by selecting Experiments in the section nav.
  • project: Your project ID. You can find these Project IDs in the Google Cloud console welcome page.
  • location: See List of available locations
  • state: Possible values for state, which shows up as "status" in the Google Cloud console, are:
    • aiplatform.gapic.Execution.State.CACHED
    • aiplatform.gapic.Execution.State.CANCELLED
    • aiplatform.gapic.Execution.State.COMPLETE
    • aiplatform.gapic.Execution.State.FAILED
    • aiplatform.gapic.Execution.State.NEW
    • aiplatform.gapic.Execution.State.RUNNING

Google Cloud console

Follow these steps to delete an experiment run.
  1. In the Google Cloud console, go to the Experiments page.
    Go to Experiments
  2. In the experiment details page, click the name of the experiment that is associated with the experiment run you want to delete. The Experiment runs page appears with the list of all the experiment runs for that experiment.
  3. Select the checkbox associated with the run that you want to delete. The Delete button appears.
  4. Click Delete
    • Alternatively, you can go to the  options menu that's in the same row as the experiment run and select delete.

View list of experiment runs and run details

The Google Cloud console provides a visualization of the data associated with these runs.

What's next