Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions models/runs/manage-runs.mdx
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,8 @@ description: Move runs between projects or teams using the W&B App.
Before you begin, ensure you have the necessary permissions to move runs between projects or teams. You must have access to the run at its current and new locations.
</Note>

Moving runs is supported only through the W&B App UI. There is no Python SDK or Public API method for moving runs between projects or teams, and W&B does not support copying runs.

To move runs from one project to another or between teams:

1. Navigate to the project that contains the runs you want to move.
Expand All @@ -19,6 +21,10 @@ To move runs from one project to another or between teams:
When you move a run, historical artifacts associated with it are not moved. To move an artifact manually, you can use the [`wandb artifact get`](/models/ref/cli/wandb-artifact/wandb-artifact-get/) SDK command or the [`Api.artifact` API](/models/ref/python/public-api/api/#artifact) to download the artifact, then use [`wandb artifact put`](/models/ref/cli/wandb-artifact/wandb-artifact-put/) or the `Api.artifact` API to upload it to the run's new location.
</Note>

## Delete runs programmatically

To delete a run from the Python Public API, use [`Run.delete()`](/models/ref/python/public-api/run#method-run-delete). This is the only run-management operation available through the Public API; moving runs between projects must be performed in the W&B App.

{/* ## Move runs to a group

You can group runs by moving them into an existing group or creating a new group.
Expand Down
19 changes: 19 additions & 0 deletions models/track/config.mdx
Original file line number Diff line number Diff line change
Expand Up @@ -255,6 +255,25 @@ To load a configuration file other than `config-defaults.yaml`, use the `--confi
python train.py --configs other-config.yaml
```

### Nested dictionaries in config

The `desc`/`value` schema used by `config-defaults.yaml` does not support nested dictionaries inline. To log a nested dictionary as config, pass it directly to `wandb.init()` in Python:

```python
config = {
"EPISODE_BUDGET": {
"cramped_room": 5000,
"coordination_ring": 5000,
"counter_circuit_o_1order": 5000,
}
}

with wandb.init(project="my_project", config=config) as run:
...
```

W&B flattens nested dictionaries using dots when displaying config in the UI and when filtering with the Public API (for example, `EPISODE_BUDGET.cramped_room`). In Python, you can still access values with normal dictionary indexing: `run.config["EPISODE_BUDGET"]["cramped_room"]`.

### Example use case for file-based configs
Suppose you have a YAML file with some metadata for the run, and then a dictionary of hyperparameters in your Python script. You can save both in the nested `config` object:

Expand Down
Loading