> ## 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.

# Text-to-speech datasets

> Create and manage the text you test with

<video controls src="https://mintcdn.com/amandalmia/eC2smNZTZjZjvk8E/videos/app-text-to-speech-datasets.mp4?fit=max&auto=format&n=eC2smNZTZjZjvk8E&q=85&s=7a683dbf38db1843c47a94fd7942c033" />

A dataset is a set of lines of text for the models to read aloud. You can run many evaluations on the same dataset.

## See your datasets

In the sidebar, click **Text-to-Speech**, then click the **Datasets** tab. Each row shows how many lines the dataset has (**Items**) and how many evaluations used it (**Evals**).

![The Datasets tab of the Text-to-Speech page, listing Tutor replies in Hindi](https://mintcdn.com/amandalmia/eC2smNZTZjZjvk8E/images/app/tts-datasets.png?fit=max\&auto=format\&n=eC2smNZTZjZjvk8E\&q=85\&s=77cb1b388a2060f8cd5dc06f34c94bff)

## Create a dataset

1. On the **Datasets** tab, click **New dataset**.
2. Under **Dataset name**, type a name, for example `Tutor replies in Hindi`, and click **Create**.
3. On the dataset page, type a line of text in the first row.
4. Click **Add another row** for each other line. To add many at once, click **Choose CSV** in the **Upload CSV** box.
5. Click **Save**.

Each line can be up to 200 characters long.

A dataset is also created when you enter text on the **Dataset** tab of a new evaluation.

![The Dataset tab of a new evaluation, with Tutor replies in Hindi typed as the dataset name and the first row filled in](https://mintcdn.com/amandalmia/eC2smNZTZjZjvk8E/images/app/tts-dataset-create.png?fit=max\&auto=format\&n=eC2smNZTZjZjvk8E\&q=85\&s=ad415e567bd3061900f56597cadf0ae9)

## Change a dataset

Click a dataset on the **Datasets** tab to open it. On the dataset page you can:

* Add lines with **Add another row** or **Choose CSV**.
* Change a line by typing in its box.
* Remove a line with the bin button on its row. The button shows once your other changes are saved, and only while the dataset has more than one line.
* Rename the dataset with the pencil button next to its name.

Click **Save** to keep your changes.

![The dataset page for Tutor replies in Hindi, with three saved rows, an empty fourth row, a pencil button next to the name and a New evaluation button](https://mintcdn.com/amandalmia/eC2smNZTZjZjvk8E/images/app/tts-dataset-edit.png?fit=max\&auto=format\&n=eC2smNZTZjZjvk8E\&q=85\&s=50bc6093828730c5e6e70d43848a9eab)

## Run an evaluation on a dataset

On the dataset page, click **New evaluation**. The dataset is already chosen, so you only pick the language and models. Save any changes first.

## Delete a dataset

On the **Datasets** tab, click the bin button at the end of the dataset's row, then click **Remove**. This cannot be undone.

![The Delete dataset window, with Cancel and Remove buttons](https://mintcdn.com/amandalmia/eC2smNZTZjZjvk8E/images/app/tts-dataset-delete.png?fit=max\&auto=format\&n=eC2smNZTZjZjvk8E\&q=85\&s=c3efbd087c4acbddee689ae23de6b792)

## Next steps

> [card] [Benchmark text-to-speech models](https://docs.calibrate.artpark.ai/quickstart/text-to-speech.md)
>
> Run an evaluation on your text

> [card] [Text-to-speech metrics](https://docs.calibrate.artpark.ai/core-concepts/text-to-speech.md)
>
> What each score means


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