Add datasets to the Tabby model card

[NielsRogge]

Hi @soCromp 🤗

I'm Niels and work as part of the open-source team at Hugging Face. I discovered your work through Hugging Face's daily papers and your model card: https://huggingface.co/sonicc/tabby-distilgpt2-diabetes.

I noticed you've already uploaded a model checkpoint to Hugging Face. It would be great to make the datasets used in your paper also available on the 🤗 Hub to improve their discoverability and visibility. We can add tags to help people find them when filtering through datasets, and link them to the model card.

The README mentions the following datasets: Diabetes, Travel, Adult, Abalone, Rainfall, and House. Would you be open to adding these datasets, with their respective URLs, to the model card? This would make your work even more accessible to the community!

Let me know if you're interested or need any guidance!

Cheers,

Niels
ML Engineer @ HF 🤗

[NamburiSrinath]

Hi @NielsRogge,

Thanks for showing interest in our work.

As different datasets might have various licenses and we just use it in our work, we are not sure if we can add

But for the downloadable links part, our Appendix B in paper link has more details regarding each dataset along with the downloadable links (Refer Table 5)

Here are the downloadable links for the datasets used in our work

• Diabetes: https://www.openml.org/search?type=data&sort=runs&id=37&status=active
• Travel: https://www.kaggle.com/datasets/tejashvi14/tour-travels-customer-churn-prediction/data
• Adult: https://archive.ics.uci.edu/dataset/2/adult
• Abalone: https://www.openml.org/search?type=data&sort=runs&id=183&status=active
• Rainfall: https://www.openml.org/search?type=data&status=active&id=41539&sort=runs
• House: https://scikit-learn.org/stable/modules/generated/sklearn.datasets.fetch_california_housing.html
• Glaucoma: https://huggingface.co/datasets/AswanthCManoj/glaucoma_diagnosis_json_analysis

Our work is still in progress and we plan to add these details to model card (or create a new HF repo) sometime later.

Once again, thanks a lot for checking our work.