Delete runs

This page shows how to delete runs from a W&B project interactively in the W&B App or programmatically with the W&B Python SDK. This page also explains the logic W&B uses to remove deleted run data from object storage on self-managed deployments. Based on your use case, select one of the following tabs to learn how to delete a run:

W&B App
Python

Navigate to the project that contains the runs you want to delete.
Select the Runs tab.
Select the checkbox next to the runs you want to delete.
Select the Delete button (trash can icon) above the table.
From the drawer that appears, select Delete.

For projects that contain many runs, use the search bar to filter runs with Regex, or use the filter button to filter runs by status, tags, or other properties.

You can delete runs programmatically with wandb.apis.public.Run.delete(). To also remove artifacts associated with the run, set delete_artifacts=True.Replace [ENTITY] with your W&B entity name and [PROJECT] with your project name.

import wandb

api = wandb.Api()
runs = api.runs("[ENTITY]/[PROJECT]")
for run in runs:
    if run.state == "finished":  # Replace with your own condition
        run.delete(delete_artifacts=False)

To remove individual files attached to a run, like logged media:

Obtain the relevant file handles with Run.files().
Use File.delete() to delete individual files.

You can’t reuse a run ID, even after you delete the run. If you try to start a new run with that ID, it fails with an error.

When you delete a run and choose to delete associated artifacts, the artifacts are permanently removed and can’t be recovered, even if the run is restored later. This includes artifacts linked to the Registry.

Run deletion flowchart

The following diagram illustrates the run deletion process, including how W&B handles associated artifacts and Registry links:

When deleted run data is removed from storage

On W&B Dedicated Cloud and W&B Self-Managed, the GORILLA_DATA_RETENTION_PERIOD environment variable controls how long W&B retains deleted run data before permanently removing it from object storage. This setting doesn’t remove artifacts. Artifacts follow the deletion and garbage collection flow described in Delete an artifact. When you set or change GORILLA_DATA_RETENTION_PERIOD, the effect is irreversible for data past the retention window. Back up your database and bucket before you enable or tighten retention. For the reference table and warnings, see Configure environment variables. Even after run or file deletion and retention processing, bucket usage can lag while background jobs catch up. W&B doesn’t guarantee immediate reclamation of object storage. For an overview of artifacts versus run data, timing expectations, and optional operator actions, see Manage bucket storage and costs.

If deletions don’t appear as expected in the W&B App when you use the Public API, upgrade the W&B Python SDK to a current release and retry.

Guides

Integrations

Reference

Run deletion flowchart

When deleted run data is removed from storage

​Run deletion flowchart

​When deleted run data is removed from storage

Run deletion flowchart

When deleted run data is removed from storage