> ## Documentation Index
> Fetch the complete documentation index at: https://docs.valkyrie.vals.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Datasets resource

> Create and manage the datasets that belong to a benchmark; tasks manages their contents.

Access these methods through `client.datasets`.

## `client.datasets.create`

Create a dataset in a benchmark, named like a benchmark.

An archived dataset keeps its name, so restore it instead of creating it again.

```python Signature theme={null}
async def create(benchmark: str, name: str, *, description: str = '') -> DatasetResponse
```

**Parameters**

benchmarkrequired · str

namerequired · str

descriptionoptional · str · default: `''`

**Returns**

DatasetResponse

For examples, see the [Python SDK guide](https://docs.valkyrie.vals.ai/sdk/resources.md).

## `client.datasets.list`

List a benchmark's datasets, adding archived ones when include\_archived is true.

```python Signature theme={null}
async def list(benchmark: str, *, include_archived: bool = False) -> DatasetsResponse
```

**Parameters**

benchmarkrequired · str

include\_archivedoptional · bool · default: `False`

**Returns**

DatasetsResponse

For examples, see the [Python SDK guide](https://docs.valkyrie.vals.ai/sdk/resources.md).

## `client.datasets.get`

Fetch one dataset by name.

```python Signature theme={null}
async def get(benchmark: str, name: str) -> DatasetResponse
```

**Parameters**

benchmarkrequired · str

namerequired · str

**Returns**

DatasetResponse

For examples, see the [Python SDK guide](https://docs.valkyrie.vals.ai/sdk/resources.md).

## `client.datasets.update`

Replace a dataset's description.

```python Signature theme={null}
async def update(benchmark: str, name: str, *, description: str) -> DatasetResponse
```

**Parameters**

benchmarkrequired · str

namerequired · str

descriptionrequired · str

**Returns**

DatasetResponse

For examples, see the [Python SDK guide](https://docs.valkyrie.vals.ai/sdk/resources.md).

## `client.datasets.remove`

Archive a dataset; restore brings it back.

```python Signature theme={null}
async def remove(benchmark: str, name: str) -> DatasetResponse
```

**Parameters**

benchmarkrequired · str

namerequired · str

**Returns**

DatasetResponse

For examples, see the [Python SDK guide](https://docs.valkyrie.vals.ai/sdk/resources.md).

## `client.datasets.restore`

Restore an archived dataset.

```python Signature theme={null}
async def restore(benchmark: str, name: str) -> DatasetResponse
```

**Parameters**

benchmarkrequired · str

namerequired · str

**Returns**

DatasetResponse

For examples, see the [Python SDK guide](https://docs.valkyrie.vals.ai/sdk/resources.md).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.