Run states ↗
noOriginal Documentation
Documentation Index#
Fetch the complete documentation index at: https://docs.wandb.ai/llms.txt Use this file to discover all available pages before exploring further.
Learn about the different states a W&B run can have.
The Run state indicates the current status of a W&B run. You can view the state of a run in the W&B App or programmatically using the W&B Python SDK.
Run states#
The following table describes the possible states a run can be in:
| State | Description |
|---|---|
Crashed | Run stopped sending heartbeats in the internal process, which can happen if the machine crashes. |
Failed | Run ended with a non-zero exit status. |
Finished | Run ended and fully synced data, or called wandb.Run.finish(). |
Killed | Run was forcibly stopped before it could finish. |
Running | Run is still running and has recently sent a heartbeat. |
Pending | Run is scheduled but not yet started (common in sweeps and Launch jobs). |
Run states in sweeps#
When runs are part of a sweep, their states behave independently from the sweep’s status:
- Individual run states reflect each run’s execution status (Running, Finished, Failed, etc.)
- Sweep status controls whether new runs are created, not how existing runs execute
- Pausing or stopping a sweep doesn’t affect already-running runs
- Only cancelling a sweep forcibly kills running runs (changes their state to
Killed)
For a detailed explanation of how sweep and run statuses interact, see Understanding sweep and run statuses.
View the state of a run#
Programmatically or interactively view a run’s state with the Python SDK or W&B App.
Use the state property of the wandb.Api.Run object to access the current state of a run.
The following code snippet retrieves and prints the state of all runs in a specified project. Copy and paste the following code snippet into your Python environment. Replace the values enclosed in angle brackets (< >) with your own values:
import wandb
api = wandb.Api()
runs = api.runs(path="<entity>/<project>")
# Access run object's properties
for run in runs:
print(f"Run: {run.name}")
print(f"Run state: {run.state}")
print()
```
You can apply different filters to retrieve runs from your projects based on different criteria. See [Filter runs](/models/runs/filter-runs) to learn more about filtering runs programmatically.
<span class="tab-end"></span>
<span class="tab-start" data-tab-title="W&B App"></span>
View the state of a run from the W\&B App:
1. Navigate to your W\&B project.
2. Select the **Workspace** or **Runs** tab from the project sidebar.
3. Search or scroll to the run you want to view.
4. Select the run to open the run overview page.
5. Choose the **Overview** tab.
Next to the **State** field, view the current state of the run.
<span class="tab-end"></span>
<span class="tab-group-end"></span>