Pages#

Home page#

Home page

The Home page is mainly a filterable, searchable and sortable list of datasets. It is the entry point to the Argilla web app and is composed of the following three components.

Dataset list#

In the center of the page you see the list of datasets available for your user account. The list consists of following columns:

  • Name: The name of the dataset, can be sorted alphabetically.

  • Workspace: A filterable column showing the workspace to which the dataset belongs.

  • Task: The task of the dataset. This is a filterable column.

  • Tags: User defined tags for the dataset.

  • Created at: When was the dataset first logged by the client.

  • Updated at: When was the dataset last modified, either via the Argilla web app or the client.

Dataset#

Dataset page

The Dataset page is the main page of the Argilla web app. From here you can access most of Argillaโ€™s features, like exploring and annotating the records of your dataset.

The page is composed of 4 major components:

Search bar#

Argillaโ€™s search bar is a powerful tool that allows you to thoroughly explore your dataset, and quickly navigate through the records. You can either fuzzy search the contents of your records, or use the more advanced query string syntax of Elasticsearch to take full advantage of Argillaโ€™s data models. You can find more information about how to use the search bar in our detailed search guide.

Filters#

The filters provide a quick and intuitive way to filter and sort your records with respect to various parameters. You can find more information about how to use the filters in our detailed filter guide.

Info button#

On the right side of the filters, the info button can be used to open a small overlay window. The aim is to provide the user with a shortcut to related information specific to the current view. This button appears when extra help is considered useful. (e.g. model explanation, weak labelingโ€ฆ).

Record cards#

The record cards are at the heart of the Dataset page and contain your data. There are three different flavors of record cards depending on the task of your dataset. All of them share the same basic structure showing the input text and a horizontal ellipsis (or โ€œkebab menuโ€) on the top right that lets you access the recordโ€™s metadata. Predictions and annotations are shown depending on the current mode and task of the dataset.

Hint

Use the menu on the bottom left corner of the page to select how many record cards you want to see per page: 1, 20, 50 or 100. This allows you to annotate record by record or in bigger bulks.

Check out our exploration and annotation guides to see how the record cards work in the different modes.

Text classification#

In this task the predictions are given as tags below the input text. They contain the label as well as a percentage score. When in Explore mode annotations are shown as tags on the right together with a symbol indicating if the predictions match the annotations or not. When in Annotate mode predictions and annotations share the same labels (annotation labels are darker).

A text classification dataset can support either single-label or multi-label classification - in other words, records are either annotated with one or multiple labels.

Note

If the record has multiple inputs, these will appear in alphabetical order.

Token classification#

In this task annotations are given as colored highlights in the input text, while predictions are indicated by underlines. At the top of the record list you will find a legend that connects the colors to the respective labels. When in Annotate mode you can remove annotations or add new ones by simply selecting the desired text.

Hint

When using the score filter, the predictions that do not fall in the selected range will be missing the solid thin line.

Text2Text#

In this task, predictions and annotations are displayed at the same level to allow for easy comparison of versions. The user interface supports the display of multiple predictions for one record, ordered by score. In Annotate mode, you can edit a prediction directly in the text box to produce your annotation.

Dataset settings#

Dataset page

You can access this page by clicking on the settings icon next to the user icon inside of the Dataset view.

In this page, you will be able to access the settings of a specific dataset and make modifications.

Description and annotation guidelines#

Soon you will be able to edit this information.

Labels#

Here you can define your labeling schema for this dataset, see the current one and add new labels to it.

Danger zone#

In this area, you will find a button to delete the current dataset.