Improve dataset card: Add metadata, paper/code links, and detailed description

#1
by nielsr HF Staff - opened

This PR significantly enhances the dataset card for FinAuditing by:

  • Adding relevant task_categories (text-classification, question-answering), language (en), and tags (finance, auditing, xbrl, llm-evaluation) to the metadata for improved discoverability.
  • Linking to the associated paper: https://huggingface.co/papers/2510.08886.
  • Providing links to the main GitHub repository (https://github.com/The-FinAI/FinAuditing.git) and the evaluation framework (https://github.com/The-FinAI/FinBen).
  • Including a detailed description of the benchmark and its subtasks (FinSM, FinRE, FinMR), extracted from the paper abstract and the GitHub README.
  • Adding the BibTeX citation for proper attribution.

These additions will make the FinAuditing dataset much more informative and accessible on the Hugging Face Hub.

Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment