Improve dataset card: Add license and tags, sync num_examples, rename usage section

#2
by nielsr HF Staff - opened

This PR enhances the dataset card by:

  • Adding license: cc-by-sa-4.0 to the YAML metadata and updating the "License" section in the Markdown content to explicitly state this license, reflecting common practice for MS MARCO derivatives on the Hub.
  • Correcting the num_examples in the dataset_info.splits metadata to accurately match the statistics provided in the Markdown content's "Key Features" and "Data Splits" table.
  • Introducing retrieval-augmented-generation and curriculum-learning as tags to the metadata, improving discoverability based on the paper's core concepts.
  • Renaming the "Usage" section to "Sample Usage" for better consistency with Hugging Face guidelines.
  • Retaining the existing badges for the paper (ArXiv), GitHub repository, and project page.
Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment