Hugging Face's logo

Datasets:
smmile
/
SMMILE

Tasks:
Image-Text-to-Text
Modalities:
Image
Text
Formats:
parquet
Languages:
English
Size:
< 1K
ArXiv:
Tags:
medical
multimodal
in-context-learning
vqa
benchmark
Libraries:
Datasets
pandas
License:
Dataset card Data Studio Files
xet
Community
2

Improve dataset card: Add metadata, links, introduction, usage, and citation

#2
by nielsr HF Staff - opened
base: refs/heads/main
←
from: refs/pr/2
Discussion Files changed
+62
-0
nielsr

This Pull Request significantly enhances the SMMILE dataset card by:

  • Adding crucial metadata: task_categories: image-text-to-text, license: cc-by-4.0, language: en, and relevant tags (medical, multimodal, in-context-learning, vqa, benchmark).
  • Populating the content section with a comprehensive introduction directly from the project's GitHub README.
  • Including prominent links to the associated Hugging Face paper (https://huggingface.co/papers/2506.21355), the project page (https://smmile-benchmark.github.io), and the GitHub repository (https://github.com/smmile-benchmark/SMMILE).
  • Providing clear sample usage instructions for loading the dataset.
  • Adding explicit license information and the BibTeX citation for proper attribution.
  • Including the project logo and acknowledgments.

These additions ensure the dataset is more discoverable, well-documented, and easier for researchers to use on the Hugging Face Hub.

 Improve dataset card: Add metadata, links, introduction, usage, and citationd2ccf653
mr02 changed pull request status to merged

Sign up or log in to comment