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

# Datasets

> Operations for datasets

## Create dataset

Create a new empty dataset. Add items separately

```bash theme={null}
calibrate datasets create [flags]
```

**Options**

| Option | Type | Description |
| - | - | - |
| `--dataset-type` | stt | **Required.** stt items carry audio + ground-truth text. `tts` items carry text only (options: stt, tts) |
| `-n, --name` | string | **Required.** Dataset name, unique within the workspace |

**Examples**

```bash theme={null}
calibrate datasets create --name <value> --dataset-type stt
```

## Get dataset

Get a dataset with all of its items

```bash theme={null}
calibrate datasets get [flags]
```

**Options**

| Option | Type | Description |
| - | - | - |
| `--dataset-id` | string | **Required.** The dataset to retrieve |

**Examples**

```bash theme={null}
calibrate datasets get --dataset-id f47ac10b-58cc-4372-a567-0e02b2c3d479
```

## Upload STT dataset item

Upload one audio clip with its transcript and append it to an STT dataset

```bash theme={null}
calibrate datasets upload-item [flags]
```

**Options**

| Option | Type | Description |
| - | - | - |
| `-a, --audio` | string | **Required.** WAV file of the spoken audio, at most 25 MB and 10 minutes long |
| `--dataset-id` | string | **Required.** The STT dataset to add the item to |
| `-t, --text` | string | **Required.** Ground-truth transcript of the audio |

**Examples**

```bash theme={null}
calibrate datasets upload-item --dataset-id f47ac10b-58cc-4372-a567-0e02b2c3d479
```


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