Skip to content

Dataset · CroviaTrust

Hugging Face training data disclosure verification log

CroviaTrust/tacet-disclosure-ledger

Tabular and text registry of hourly Hugging Face model card checks for training data disclosures.

What's inside

Contains 100K to 1M English records in tabular and text format. Every hour, the Crovia system downloads public model cards for a rotating list of Hugging Face models, checks for training data information, and cryptographically signs the result. Observations have been conducted since September 19, 2026.

What you can do with it

  • Transparency auditing and openness analysis of Hugging Face models
  • Training classifiers to automatically detect dataset mentions in documentation
Languages
English
Size
100K–1M rows
Downloads this month
29.9K
Likes
0

How to load it

from datasets import load_dataset

ds = load_dataset("CroviaTrust/tacet-disclosure-ledger")