FIUND · FREE DATASET TOOLS

Dataset Card Builder

Give buyers and collaborators a clear description of your collection. Build structured documentation without inventing missing facts or assigning permissions you have not verified.

Free · No signup · Files and text are processed in this browser. Tool inputs are not uploaded or added to enquiry links. Standard site analytics measure page visits and tool actions only. Export your work before leaving; inputs are not saved.

Describe what is known. Leave unknown fields blank; the export will mark them “Not documented”. No license or permission is inferred. Language codes are included in README metadata; license status remains in the body for review.

Dataset name (required)Description and intended taskModalities and file formatsLanguages (comma-separated language codes, if known)Size and units (files, hours, rows; measured or estimated)Source and collection methodCollection dates, equipment and environmentsCoverage and participant selectionAnnotations and quality checksTrain, validation and test split methodOwnership, consent and permissions evidenceLicense identifier or commercial terms statusRestrictions and prohibited usesPersonal information and access controlsKnown limitations, bias and missing coverageMaintainer and contact details

How to use this tool

Describe the dataset, source, collection process and intended task.
Record evidence and limitations; leave unknown fields blank.
Generate and download a README, JSON card or CSV datasheet.

Questions and limitations

Does the builder assign a license?

No. It records the license or terms status you enter in the document body. It does not insert an unverified license into Hugging Face metadata or establish licensing authority.

Can I use the README on Hugging Face?

The export is Markdown with YAML front matter. Valid language codes are included as metadata. Review the README and add any task-specific Hub metadata you need before publishing.

Practical guides

← All free dataset tools