Argilla 2.4: Easily Build Fine-Tuning and Evaluation Datasets on the Hub — No Code Required

We are incredibly excited to share the most impactful feature since Argilla joined Hugging Face: you can prepare your AI datasets without any code, getting started from any Hub dataset! Using Argilla’s UI, you can easily import a dataset from the Hugging Face Hub, define questions, and start collecting human feedback.

Not familiar with Argilla? Argilla is a free, open-source data-centric tool. Using Argilla, AI developers and domain experts can collaborate and build high-quality datasets. Argilla is part of the Hugging Face family and fully integrated with the Hub. Want to know more? Here’s an intro blog post .

This new feature democratizes building high-quality datasets on the Hub:

First, you need to deploy Argilla. The recommended way is to deploy on Spaces following this guide . The default deployment comes with Hugging Face OAuth enabled, meaning your Space will be open for annotation contributions from any Hub user. OAuth is perfect for use cases when you want the community to contribute to your dataset. If you want to restrict annotation to you and other collaborators, check this guide for additional configuration options.

This new workflow streamlines the import of datasets from the Hub, but you can still import datasets using Argilla’s Python SDK if you need further customization.

In this first version, the Hub dataset must be public. If you are interested in support for private datasets, we’d love to hear from you on GitHub .

Once Argilla is running, sign in and click the “Import dataset from Hugging Face” button on the Home page. You can start with one of our example datasets or input the repo id of the dataset you want to use.

Argilla automatically suggests an initial configuration based on the dataset’s features, so you don’t need to start from scratch, but you can add questions or remove unnecessary fields. Fields should include the data you want feedback on, like text, chats, or images. Questions are the feedback you wish to collect, like labels, ratings, rankings, or text. All changes are shown in real time, so you can get a clear idea of the Argilla dataset you’re configuring.

Once you’re happy with the result, click “Create dataset” to import the dataset with your configuration. Now you’re ready to give feedback!